REPLACEMENT OF HAZARDOUS MATERIAL IN WIDE WEB
           FLEXOGRAPHIC PRINTING PROCESS
                           by

                       Paul B. Kranz
      Erie County Department of Environment and Planning
                    Buffalo, NY  14202

                 Thomas R. Williamson, HI
                 Recra Environmental, Inc.
                    10 Hazelwood Drive
                    Amherst, NY 14228

                           and
                      Paul M. Randall
                      CR-816762-02-0
                      Project Officer

                     Paul M. Randall
             Pollution Prevention Research Branch
            Risk Reduction Engineering Laboratory
                   Cincinnati, OH  45268
     RISK REDUCTION ENGINEERING LABORATORY
       OFFICE OF RESEARCH AND DEVELOPMENT
      U.S. ENVIRONMENTAL PROTECTION AGENCY
                 CINCINNATI, OH 45268

-------
                                             NOTICE

        The information in this document has been funded wholly or in part by the United States Environmental
Protection Agency under CR-816762-02-0 to Erie County Department of Environment and Planning. It has been
subjected to the Agency's peer and administrative review, and it has been approved for publication as an EPA
document. Mention of trade names of commercial products does not constitute endorsement or recommendation
for use.

-------
                                            FOREWORD

        Today's rapidly developing and changing technologies and industrial products and practices frequently
carry with them the increased generation of materials that, if improperly dealt with, can threaten both public
health and the environment.  The U.S. Environmental Protection Agency (EPA) is charged by Congress with
protecting the Nation's land, air, and water resources.  Under a mandate of national environmental laws, the
agency strives to formulate and implement actions leading to a compatible balance between human activities and
the ability of natural systems to support and nurture life.  These laws direct the EPA to perform research to
define our environmental problems, measure the impacts, and search for solutions.

        The Risk Reduction Engineering Laboratory is responsible for planning, implementing, and managing
research, development, and demonstration programs to provide an authoritative, defensible engineering basis in
support  of the  policies, programs, and regulations  of the EPA with respect to drinking water, wastewater,
pesticides, toxic substances, solid and hazardous wastes, Superfund-related activities, and pollution prevention.
This publication is one of the products that research and provides a vital communication link between the
researcher and the user community.

        This report describes the results of a technical and economic evaluation of the substitution of hazardous
materials in the production of flexible packaging utilizing a wide-web flexographic printing process. The objective
of the study was to substantiate the reduction of emissions of volatile organic compounds through the use of
water based inks in lieu of solvent based inks.
                                     E. Timothy Oppelt, Director
                                Risk Reduction Engineering Laboratory
                                                 111

-------
                                             ABSTRACT

        This report describes the technical and economic evaluation of the  substitution of nonhazardous
materials for the manufacture of flexible packaging in a wide web flexographic printing process. The objective
of this study was to evaluate, on a technical and economic basis, the impact of the substitution of'water-based
inks for solvent-based inks in a flexographic printing process.

        In order to reduce Volatile Organic Compound (VOC) emissions by switching from the use of solvent-
based inks to water-based inks, several equipment modifications and a feedstock substitution were completed.
Water-based inks containing, on average, 72.5% percent less VOC were utilized in lieu of, and in conjunction
with, traditional solvent-based inks. Equipment modifications included dryer capacity enhancement, press roller
modification, ink handling equipment upgrade and the installation of an in-line corona treatment system.

        The ink substitution resulted in a reduction hi the emissions generated from the printing process.  For
each percent increase in water-based ink utilization, a reduction in VOC emissions of 14 pounds was calculated.
This was based on a usage rate of about 2250 pounds of ink per week which caused a VOC emission of about
1570 pounds. The substitution, typically, did not adversely affect product quality or nonhazardous scrap waste
generation.  An average reduction of 95% of liquid F003 waste from waste ink  and-deaning solvents was
recorded during the study period, but was ascertained to be due to operational practice changes. ;

        To  complete the economic evaluation,  the costs of press modifications, ancillary equipment, waste
disposal, inks and solvent were obtained. A payback period and project net present value were calculated.

        The project has a positive net present value of $39,165 and a payback period of 2.5 years, based on 21%
utilization of water-based ink.  If full conversion to water-based inks is implemented, the payback period is
reduced to approximately 0.5 years ($62,901/$117,078).                                      |

        Additional benefits resulting from the reduction of VOC emissions and liquid hazardous waste have been
an improved working environment from reduced indoor air pollutants, reduced handling of hazardous solvents
by employees, and the appreciation by company employees of the need to make a conscious effort to further
reduce waste generation.

        This report was submitted in fulfillment of CR-816762-02-0 by Erie County Department of Environment
and Planning under the sponsorship of the U.S. Environmental Protection Agency. This report covers a period
from September 1990 to September 1991, and work was completed as of April 1992.
                                                 IV

-------
                                    TABLE OF CONTENTS
                                                                                        Page
NOTICE 	;. . v-;-•	kj. .^	                  y
FOREWORD	'"; ,'T.	 . f! . .	'.'.',	\	  iii
ABSTRACT	'.'.'.'.'.'.'.''.'.'.'.'.'.'.'.'.'.'.	  iv
LIST OF FIGURES	  	'	  yj
LIST OF TABLES	 ..I.  ..      vii
ACKNOWLEDGEMENT	'.'.'.'.'.','.'.'. viii

I.      INTRODUCTION	  i
       Program Overview	_	  -±
       Project Objectives	  1
       Project Description	  i
       Industrial Participant	  2
       Regulatory History	  2

II.     CONCLUSIONS AND RECOMMENDATIONS	  4
       Effect of Water-Based Ink Use on VOC Emissions	:	  4
       Liquid-Phase Solid Waste	  5
       Non-Hazardous 'Scrap Waste'	;	  5
       Corona Treatment	'.	  5
       Ink Performance	  6
       Project Summary	  g

III.    TECHNOLOGY DESCRIPTION	  7
       Flexographic Printing Process	  7
       Critical Surface Tension	  9
       Other Considerations	.	n

IV.    DESCRIPTION OF TECHNICAL METHODS (EXPERIMENTAL)  	  13
       Waste Minimization Methodology	  13
       Technical Approach	,	  15
       Technical Evaluation	15

V.     RESULTS	23
       Historical Data  	23
       Operating Data	25
       Analytical Results on Solid Wastes	<	25
       Product Performance	i	28

VI.    DATA SUMMARY AND ANALYSIS	30
       Historical Data		30
       Operating Data Results and Discussion	30
       Economic Analysis	35

VII.    REFERENCES	44

VIII.   BIBLIOGRAPHY	,	,	45

IX.    APPENDICES
       A     Sample Ink Calculations	46
       B     Economic Calculations	61
       C     Critical Surface Tension	65
       D     Ink Durability Tests		67

-------
                                        LIST OF FIGURES




                                                                                              Pagg



Figure 1. - Transfer of Ink to Substrate in Flexographic Printing	   8




Figure 2. - Wide Web Flexographic Printing Process Flow Diagram	i	  18




Figure 3. - Historical Data: Monthly Results (1990) 	25




Figure 4. - WRITE Data:  Weekly Results (1991)	27




Figure 5. - VOC Emissions: Historical and WRITE Data	31




Figure 6. - VOC Reduction: Through Water-Based Ink Use	33
                                                VI

-------
                                       LIST OF TABLES




                                                       i                                  Page



Table 1. - Historical Data	23




Table 2. - WRITE Data	25




Table 3. - Ink Waste Composition	27




Table 4. - VOC Reduction	34




Table 5. - Economic Factors Reviewed	36




Table 6. - Fixed Costs	36




Table 7. - Payback Period	41




Table 8. - Net Present Value	42
                                              VII

-------
                                     ACKNOWLEDGEMENT

       The evaluation of this waste minimization technology was made possible through the cooperation of
Lustreprint Company, Division of Mark IV Industries, located at 622 Northumberland Street, Buffalo, New York.
Our thanks and appreciation is extended to Mr. Dana J. Morgan, Lustreprint General Manager, without whose
assistance this  project would not be possible.  The work was performed jointly by the USEPA, Erie County
Division of Environmental Compliance  and Recra Environmental, Inc., Amherst, New York, and 'reviewed by
the New York Center for Hazardous Waste Management in Amherst, New York, and by Fred Shapiro of Pro-
Flex Consultants and Dr. William Shaeffer of the Graphic Arts Technical Foundation.
                                               viu

-------
                                            SECTION 1

                                         INTRODUCTION

PROGRAM OVERVIEW

        This  report is one of five innovative waste minimization technology evaluations which are to be
conducted under  the cooperative  agreement program between the U.S.  Environmental Protection Agency
Pollution Prevention Research Branch's Risk Reduction Engineering Laboratory and Erie County, entitled the
"Waste Reduction Innovative Technology Evaluation" (WRITE) Program, Contract #CR-816762-02-0.  This
project is a technical and economic review of the efforts of a wide web flexographic printing firm to substitute
water-based inks for solvent-based inks in the manufacture of flexible packaging using plastic sheet substrates.
(The term *water-based' inks includes inks that have a low solvent content—usually less than 25%.)

        The initial work plan included  an evaluation of water-based adhesives as substitutes for solvent-based
adhesives in a lamination process. However, due to the well advanced state-of-the-art in water-based adhesives,
Lustreprint decided that there was no risk to a complete and immediate changeover to this technology, which
was accomplished prior to the inception of the study program.

        The project was completed under the terms of the Erie County/EPA WRITE Program as a joint effort
by Lustreprint  Company;  Erie County  Environmental  Compliance Services, Buffalo, New York; Recra
Environmental, Inc., Amherst, New York and the EPA Office of Research and Development, Cincinnati, Ohio.

PROJECT OBJECTIVES

        The objectives of the project were to evaluate the technical feasibility (particularly as related to process
implementation and product performance), the economic effect and the resulting change in volatile organic
compound  (VOC) emissions achieved  by the substitution of water-based inks for solvent-based  inks in the
printing of flexible packaging materials.  The technical evaluation objective was to quantify the reduction hi both
volatile and liquid-phase solid  hazardous wastes generated, through minimizing solvent-based material usage.

PROJECT DESCRIPTION

        This project was a study of the effectiveness and applicability of the substitution of inks to achieve waste
reduction in the wide web (greater than 16 inches wide) flexographic printing process. These studies were
carried, out to ascertain the potential for widespread implementation of such material substitution as a waste

-------
minimization technique for the wide web printing industry. This flexographic process is used to print flexible
packaging material used for a variety of products. (For example, one such product is the plastic bags used to
hold loaves of bread. This packaging typically has the name of the bakery and other information printed on it
in various colors.  Such printing can be accomplished by the flexographic process.)             ;

INDUSTRIAL PARTICIPANT
                                                   I                                    .
Company Background

        The industrial participant was Lustreprint Company, a Division of Mark IV Industries. liustreprint is
located at 622 Northumberland Avenue, hi Buffalo, New York. Lustreprint is a printer of flexible packaging.
The company's products are used in the food and snack industry, medical, industrial and consumer applications.
Manufacturing is conducted daily on a three shift basis.  The company employs 40 people at its Buffalo location.

        Printing is  completed on a number  of different web materials.  Materials  commonly used are:
Polypropylene (acrylic coated, saran coated,  and uncoated,  corona pretreated), cellophane  (safan coated),
polyester (both metallized and unmetaUized), polyethylene, and nylon (both saran coated and uncoated),.

Company's Goals

        The goals of the company are to reduce releases of all regulated volatile organic air emissions to an
extent that would eliminate the need for costly air abatement and permitting and to eliminate the generation of
all liquid-phase solid waste, characterized as hazardous waste, at the facility.

REGULATORY HISTORY

        At the time of this study, Lustreprint utilized one Hudson/Sharp 48 inch Central Impression six: color
flexo press, one Heinrich (W&H) five color flexo stack press, one laminator and two slitters. The company
applied and received approval for a permit for ah- emissions for its two printing presses from the New York State
Department of Environmental Conservation (NYSDEC) in 1974.  A subsequent permit was issued by the
NYSDEC in 1979 for emissions associated with the operation of the laminator on-site. The VOC emissions from
the printing operation are regulated by three parts of NYSDEC Environmental Conservation Law 6 N.Y.C.R.R.
Part 212 (General Process Emission Sources) regulates the construction or operating of process emission sources
and emission standards.  The printing process is regulated under Part 234 (Graphic Arts), and the printing and
lamination processes under Part 228 (Surface Coating Processes).  The applicability of Parts 228|and 234 was
triggered by the implementation of a three shift, seven day-a-week work schedule at Lustreprint in 1989. The

-------
increased activity resulted in calculated total plant emissions exceeding the baseline criterion of 100 tons per year
of volatile organic compounds.

        Parts 228  and 234 require that a facility develop a plan and implement controls to reduce overall plant
emissions to within the compliance level of 100 tons per year. As an option, Lustreprint chose to develop a
facility-wide plan to reduce emissions by eliminating the use of solvent-based inks and adhesives.  The first step
was the elimination of solvent-based adhesives in laminating. This was to be followed by a phase-in of water-
based inks to replace the existing solvent-based inks in the printing operation and is the subject of the present
study.

-------
                                             SECTION 2

                             CONCLUSIONS AND RECOMMENDATIONS                 !

        The objective of this study was to evaluate the impact of the substitution of water-based inks for solvent-
based inks in a flexographic printing process.  The intent of the substitution was to decrease the total VOC
emissions from the facility and reduce the volume of liquid hazardous waste generation.  In addition to the VOC
emissions, several other aspects of Lustreprint operations and production were impacted by the change.

EFFECT OF WATER-BASED INK USE ON VOC EMISSIONS

        As has been determined by  the study, Lustreprint experienced a  decrease  of 14 pounds of VOC
emissions per  unit percentage of  water-based ink used (from a base of about 2250  pounds of ink used-1570
pounds of VOC emissions).  Beyond the reduction of VOC emissions, the changeover to water inks has had an
impact on operations and production at Lustreprint. Mechanical modifications to the press were completed to
minimize the differences of water-based ink use. Increases in dryer  capacity  have allowed Lustreprint to print
with water-based inks with no significant loss in printing speed.  Water-based inks  were noted as being  less
forgiving with respect to imperfections in the press cylinders, printing plates, printing drum, etc. This required
closer monitoring of the printing operation by press operators resulting in improved product quality. The need
for well cleaned  cylinders  and plates required  a more involved and time consuming  cleaning procedure while
using water-based inks, increasing downtime between job or color changes.

        During the tune of the study, Lustreprint experimented with a more advanced type of dot printing. This
process uses four basic colors to simulate all other colors hi a fashion similar  to color photography.  By varying
the size of the  dots for a particular color,  in combination with the other three, .various colors could be achieved.
The use of water inks for this type of printing was beneficial. Water inks, due to the decreased wetting ability,
allowed better control of dot size, which is critical for this type of printing.

        To date, Lustreprint has  not yet  experimented with all available ink  technology. Trials with different
inks, for  different surfaces, for different printing applications continue.  Initial testing with  total water inks, as
opposed  to low solvent inks, indicate further VOC reduction may be possible.  The reduction in VOC emissions,
the reduced handling of hazardous materials by employees, and unproved working environment all support the
continued use  of water inks.

-------
LIQUID-PHASE SOLID WASTE

        In general, solid-phase waste produced at Lustreprint is non-hazardous.  Thus the focus on waste
reduction has primarily been on liquid-phase wastes, described as 'solid waste' under the Resource Conservation
and Recovery Act (RCRA).  For ease of use, the term 'solid waste' will be used with the RCRA meaning to
describe the liquid hazardous wastes generated in this project. Where solid material is produced as scrap, it shall
be differentiated by noting that it is 'scrap'.

        Lustreprint intends  to continue tracking of wastes to assist in limiting solid waste generation. In the
future, the cost of waste generation from solvent ink use might be attributed to a particular job and factored into
the pricing structure, reducing overhead costs from waste disposal. A study period of greater than one week
would be required to determine the ultimate impact of water inks on solid waste generation.

        Another facet of solid waste generation at Lustreprint, which is not directly related to ink use, is the ink
pan cleaning station. The cleaning station is. an enclosed tank which holds approximately 55 gallons of solvent.
It is used for the cleaning of large items or particularity difficult cleaning jobs. The solvent  is changed only as
needed, approximately every other month. A sample of spent solvent from this tank revealed that the material,
at the time of disposal, is approximately 50% solvent.

        Recommendations to Lustreprint for further reduction of solid waste generation included continued
waste tracking, the purchase of an ink splitter for reduction of both solvent and water waste  and alternatives to
the ink pan cleaning station such as sodium bicarbonate blast cleaning. Teflon coated ink pans provide another
opportunity to assist in reducing solid waste generation from press cleaning.

NONHAZARDOUS 'SCRAP' WASTE

        Nonhazardous waste is generated by Lustreprint in the form of scrap printed plastic sheeting which is
a result of off-spec product or quality assurance review.  The majority of scrap plastic sheeting is generated at
the time of start-up. After the plastic sheet material is loaded into the press along with the necessary inks, a
sample of the job is run. The sample is checked for a number of quality assurance parameters.  Iterations are
continued with adjustments  to the press roller  pressures and run  speed until a satisfactory print is achieved.

        It was anticipated  that the use of water-based  inks  might result in an increase in scrap generation.
Water-based inks, being less forgiving to roller or plate imperfections, might initially require additional quality
assurance checks.  Scrap totals reported by Lustreprint showed an increase from 9.9% to 14% over the course
of the last year.  However Lustreprint attributes the increase to factors other than water-based ink use, such as

-------
operator training.  Even after the implementation of all water inks is achieved, the scrap rate is expected to
return to normal and no future increase in the generation of scrap is anticipated.

CORONA TREATMENT

        Overall, the in-line corona treatment was found to assist in most printing applications.  This includes
both water- and solvent-based ink. The benefit to the printing quality centered around ink adhesion.  Lustreprint
is in the process of purchasing and installing a second corona treatment unit on its new five color stack press.

INK PERFORMANCE                                                                   :

        Although there were occasional problems, water inks were successfully used. Some problems arose after
use by the customer.  Heat and stress of the printed package material, caused by the package folding and sealing
process, at times resulted in a loss of ink adhesion.  Lustreprint stated that such losses were minimal.

        Some combinations of water and solvent inks were found to be incompatible. In most cases, depending
upon the surface printed, there was no difference noted with the use of water-based inks.

PROJECT SUMMARY

        The results of the technical evaluation for the substitution of water inks for solvent inks for flexographic
printing indicate that a reduction in VOC emissions is possible without typically compromising product quality.
The economic evaluation supports the cost effectiveness of the substitution.            •       !

        The economic analysis indicates a payback of 2.56 years for the equipment necessary to convert to water-
based inks. Using an NPV calculation, a positive cash flow of $39,165 would be achieved. If Lustreprint should
split their solid waste streams, the resulting payback, including additional equipment, drops to 2.03 years.

        The installation of the corona treater along with enhanced dryer capacity facilitated the use of water-
based inks. The substitution did not increase nonhazardous scrap waste generation, while further reductions in
the generation of liquid-phase solid hazardous waste are possible.

        This technique should be considered by any company involved in wide web flexographic printing as a
method for source reduction of VOCs through substitution of raw materials.

-------
                                             SECTION 3

                                   TECHNOLOGY DESCRIPTION

FLEXOGRAPHIC PRINTING PROCESS

Existing Process Equipment

        Flexographic printing is defined as a printing process "that uses a flexible plastic or rubber plate in a
rotary web press."1 It "is used primarily for packaging, such as plastic wrappers, corrugated boxes, milk cartons,
shower curtains, foil and paper bags."1  This process uses printing plates which are flexible and usually made
from elastomeric materials, attached to a roller or cylinder for ink application. A flexographic printing press may
contain one or several printing stations, each made up with a series of rolls (See Figure 1).

        The traditional flexographic printing station is comprised of four basic rolls.   The rubber surface
Fountain or Ink Roll is the first and acquires and transports ink for printing from the ink pan.  A metal or
ceramic-surfaced Metering Roll (Anilox Roll) follows, which is specially engraved to take a uniform amount of
ink oveir the surface area from the Fountain Roll. The Metering Roll delivers the ink to the printing plates which
are mounted by pressure sensitive "stickyback" on the Plate Cylinder. The substrate to be printed passes between
the Plate Cylinder and the metal Impression Cylinder, which applies a preset pressure against the Plate Cylinder
to impress the  ink onto the web material It  is at  this point where the actual printing occurs.  Numerous
variations  to the number and use of rolls and cylinders are common for various other printing operations.
Multiple printing stations are used for multi-color printing,  one for each ink color.

       Referring to Figure 1, the printing press itself is made up of five basic sections:

                a)      UNWIND—The material to be printed upon is fed into the press from the unwind
                       section.

               b)      TREATER—The corona discharge treater is installed between the unwinding roll and
                       the in-feed section of the printing press.

               c)      PRINTING SECnON-The material is then fed into the printing station(s).

               d)      AFTER DRYER-Next, the printed web proceeds through the overhead drying section.

-------
 o  S
O  £
                                                                                                      O)
                                                                                                      c
                                                                                                      Q.
                                                                                                      C3)

                                                                                                      O
                                                                                                      X
                                                                                                     1
                                                                                                      w
                                                                                                     J2

                                                                                                      U)
                                                                                                      c

                                                                                                     •5
                                                                                                      v_
                                                                                                      CD

                                              8

-------
                e)      REWIND—Finally, finished product (printed and dried) proceeds to the rewind section.

        Depending on the type of printing press, three basic configurations of printing stations are common:

                •       A stack type of printing press places printing stations mounted one above the other for
                        multiple color printing.

                •       Similarly, an in-line press places the stations in a tandem series.

                •       The central impression cylinder printing press is unique in that  one large impression
                        cylinder is  common to several printing stations arrayed about the perimeter of this
                        Central Impression Cylinder.

        Lustreprint utilizes a six color central impression wide web press.  Wide web printing can be used for
newspapers and wall papers.  However, Lustreprint's product is used in packaging, another major use.

Process Modifications

        The substitution of solvent-based materials by water-based inks required press modification.  The most
significant retrofit was the installation of a Enercon Corona Discharge Treater.

        Modifications of the Hudson/Sharp 48 inch Central Impression six color flexographic printing press
included the upgrade of the drying capacities, utilizing enlarged exhaust and supply fans. Additional duct work
was required along with noise abatement equipment. Metering rolls were replaced in order to apply a thinner
layer of ink to facilitate drying.  Pumps were also replaced to accommodate the new printing inks.

        The cost involved in the retrofit of one of the two on-site presses, the Heinrich (W & H) five color flexo
stack press, makes modification of that press for water-based ink use prohibitive. Future plans for the facility
would include the replacement of this press with one which could accommodate the ancillary equipment required
for water-based ink usage.

CRITICAL SURFACE TENSION

        The printing of flexible packaging using plastic film substrates is dependent on the surface characteristics
of the film material (mainly the Critical Surface Tension (CST) of the plastic film) and the wetting ability of the

-------
ink used.  Wetting ability is the property of a liquid to spread over and "wet" a surface and is dependent on the
relative surface tensions of the liquid and the printed substrate.

        Ink wetting and adhesion are enhanced by an increase in the CST of the plastic substrate.  While
theoretically the CST of printing and laminating surfaces needs to be equal  to the surface tension of the ink,
from a practical standpoint, for optimum performance, printing and laminating surfaces need a Critical Surface
Tension level about 10 dynes/cm higher than the ink to acquire the proper adhesion to the film surface* (See
also Appendix C).

        The CST for various web materials used for flexographic printing would normally be about 45 dynes/cm.
The CST for plastic materials is affected by the chemical composition of the material, slip additives which are
often used in the production of the material to  assist in the unwinding of the  rolls,  and corona treatment. The
slip additives used in the manufacture of the films have a tendency to "bloom", or migrate from within the body
mass of the substrate to the surface of the web, and cause a decrease in the film CST.  This is due to  the
formation on the surface of the web of a weakly bonded layer of slip agent. This slip agent in turn has a lower
surface tension and causes a decrease in the apparent surface tension of the web, well  below the normal 45
dynes/cm to about  30 dynes/cm.

        The impact of chemical characteristics is a result of the polarity of the portion of the molecule which
is exposed at the surface of the film.  The impact  is best  described in an article by Baler et al3.  Corona
treatment of the film  tends  to both increase  the number  of polar groups at the surface  of the film, and
                                                                                          i
temporarily retard the bloom of slip agents (it also removes slip agent on the surface of the film). This process
results in an increase in surface energy, leading to a higher CST of above 40 dynes/cm, which in turn increases
the "wettability" of the film.

        In order to increase the CST of the plastic film, the  manufacture of films, used as web material in
flexographic printing, usually includes a corona treatment step prior to completion.  However, the effects of the
film manufacturer's corona treatment will degrade over time, as surface energy declines, resulting in poorer ink
wetting and, hence, reduced ink adhesion.  To utilize the full benefit of the corona treatment, Lustreprint has
installed an in-line corona treatment device (Enercon Industries Corona Discharge Treater) subsequent to the
unwind/in-feed section of the 48 inch Hudson/Sharp Central Impression press.

        Because solvent-based inks have been traditionally used by the printing industry, there has generally been
no problem with wetting.  Solvent-based inks have a lower surface tension than the CST of the ;plastic webs
typically used.  These inks' ability to wet and adhere to the web material, coupled with their quick drying nature,
has made them the preferred choice. In order to convert to water-based inks, something must be done to assist

                                                  10

-------
in lowering their surface tension to complement the increase in the surface tension of the web achieved through,
corona treatment.
                                        > s                '   .-,*?

        The printing ability of water-based inks may be enhanced with additions to the ink formulations. The
surface tension of water is about 72 dynes/cm compared to solvent materials used in ink formulations which have
surface tensions of between 20-25 dynes/cm. To reduce the surface tension of water-based inks, small amounts
of solvent, usually alcohol, are added to the ink formulations.  A 20% by weight ethyl alcohol solution has a
surface tension of 38 dynes, sufficiently low for printing on plastic web material which, after corona treatment,
has a CST value of about 40 dynes.  For this reason, water-based inks are actually low solvent alternatives
compared to traditional solvent-based inks which have a solvent content of around 70-75%. This relatively small
amount of solvent material not only increases the wetting ability of the ink, but also assists in dissolving the slip
additives present on the surface of the web material, something that water alone cannot accomplish.

OTHER CONSIDERATIONS

        Other problems must be resolved with the use of water inks, particularly ink metering and ink drying.

Metering of Ink

        Ink film thickness  is determined hi part by the type of metering roll used. The meter roll is a finely
engraved cylinder with minute pockets or cells which receive the ink from the ink roll and deliver it to the flexo
printing plates.  The amount of ink delivered to the plates can be reduced by utilizing a meter roll engraved with
smaller volume cells.  The optimum density of cells varies with the type of printing application. However, this
cell density cannot be arbitrarily changed. This is because water-based inks have a higher solids content and any
flow cessation, e.g. for quality control samples to be cut and removed, these solids settle and tend to quickly foul
the engraved surface, producing unsatisfactory results. Thus the  optimum cell density and configuration must
be chosen to minimize such difficulties.

        Makrauer has noted that traditional chrome rolls which employ pyramid or truncated pyramid shaped
cells have better ink release properties, but that engraved ceramic meter rolls are more often used for water-
based inks, because the ceramic rolls tend to wear better, extending the life span of the meter rolls4. (Lustreprint
does not use a ceramic roll, but now has one in-house for future testing.)

        Further control of the ink  film thickness can be achieved through the use of doctor blades.  Doctor
blades are installed on the ink roll to more precisely control the amount of ink delivered to the meter rolL This
allows for maximum control of ink thickness.
                                                  11

-------
Ink Drying

        Ink drying is accomplished after each ink color application and subsequent to the completion of printing.
Ink drying speed is a function of both film thickness and dryer capacity.  These parameters must be monitored
to maintain a sufficient run speed for productivity of the press operations and to avoid web damage from high
temperature dryer air.                                                                      \

        The step used to enhance the drying of water-based ink is an increase in the dryer capacity of the press.
Dryer capacity is  a function of air volume, air velocity and air temperature. Studies have indicated4 that air
volume  and air velocity are of more importance than air temperature.  An upgrade in blower; capacity is
necessary to accomplish an increase in these, and to  reach a point where water-based inks  could be  used
interchangeably with solvent-based inks.

        The two supply fans and exhaust fan for the Hudson/Sharp Press were upgraded to obtain the necessary
drying capacity.  The two supply fans are now rated at 2690 Cubic Feet per Minute (CFM) at 10" Standard
Pressure (SP) and 3660 CFM at 13" SP.  The exhaust fan has been upgraded to 8,000 CFM at 10" SP. In
addition, a  sound  deadener and additional ductwork were required for the installation.
                                                 12

-------
                                           SECTION 4

                   DESCRIPTION OF TECHNICAL METHODS (EXPERIMENTAL)

WASTE MINIMIZATION METHODOLOGY
        There are three major sources of volatile emissions at Lustreprint, which are the laminator and two
printing presses. For purposes of tracking, monitoring and reporting emissions from the facility, the two printing
presses are treated as a combined unit and the emissions are calculated from total ink usage at the plant.  The
reporting of emissions from the plant has been initiated as part of a program which began in February 1990.
These emissions were used as a baseline and for guidance in determining the reduction of emissions from the
plant as a result of the partial changeover to water-based ink.

        For calculating VOC emissions (in pounds), a relationship between the volume of ink used (hi gallons),
its various VOC source components and the VOC content of these components is needed.  Calculation of the
quantity of VOC emissions (in pounds) is obtained by inference from 6 N.Y.CJR..R. Part 2343(e)(l)(ii) as

        VOC           =      V * (Vv)a * d(VOC)a                         Equation 1
where,
        V              =      volume of ink actually used (in gallons),
        (Vv)a          =      volume fraction VOC of ink (minus water and excluded VOC), as applied, and
        d(VOC)a       =      density of VOC in ink  as applied (minus water and excluded VOC), (hi
                              pounds per gallon).
        From 6 N.Y.C.R.R Part 228.2(b)(2), we find
        (Vn)a          =      1 - (VOC)a/d(VOC)a                         Equation 2
where,
        (Vn)a          =      volume fraction of solids (labelled as (Vs)a hi §228.2(b)(2)), and
        (VOC)a        =      the VOC content of the coating, at application, expressed as a mass of VOC
                              (in pounds),  per volume of coating (hi gallons), minus water and excluded
                              VOC.
        Solving Equation 2 for d(VOC)a, we obtain d(VOC)a  = (VOC)a/(l-(Vn)a), and since (Vn)a tends to
zero as the volume fraction of pigment solids is quite small

        d(VOC)a       =      (VOC)a.                                      Equations

                                               13

-------
         Substituting Equation 3 into Equation 1, we obtain VOC  = V * (Vv)a * (VOC)a.  (VOC)a can be
 determined from the equation found in 6 N.Y.C.R.R §228.2(b)(2):
         (VOC)a        =      (DcW(Wv)a - fWw)a - (We)a)                 Equation 4
                               1 - ((Vw)a + (Ve)a)
 where,
         (Dc)a          =      coating density, as applied (in pounds per gallon),
         (Wv)a          =      weight fraction of total volatiles in the coating, as applied,
         (Ww)a          =      weight fraction of water in the coating, as applied,
         (We)a          =      weight fraction excluded VOCs hi the coating, as applied,      ;
         (Vw)a          =      volume fraction water of ink, as applied, and
         (Ve)a          =      volume fraction excluded VOC of ink, as applied.
         For the solvent-based ink, (Vw)a and (Ww)a = 0. In addition, Lustreprint uses no excluded VOCs and
 the terms (Ve)a and (We)a are also = 0.  Reducing Equation 4, we obtain (VOC)a = (Dc)a(Wv)a.

         (Dc)a, which has a value  of 8.485, is an average value for ink density previously determined by
 Lustreprint and is used in reporting to New York State. (Wv)a of 0.635 is also an average value used in such
 reporting.  Thus

        (VOC)a         =      8.485*0.635
                               5.388 lbs./gal.

        Substituting this value into Equation 1, we get

        VOC           =      V *  (Vv)a * 5388.

        The volume of ink, V (in gallons), is obtained from  operating records. (Vv)a may be found according
 to 6 N.Y.C.R.R §2343(a)(l)(i) by

        (Vv)a          =       1 - ((Vw)a +  (Ve)a + (Vn)a)
where,                                            ]
        (Vn)a          =      volume fraction solids of ink, as applied.

        For the solvent-based ink, (Vw)a  = 0. Lustreprint does not use excluded VOCs and the term (Ve)a
is also = 0.  This equation therefore reduces to (Vv)a = 1 - (Vn)a.

                                                14 !

-------
         Since the volume fraction of pigment solids is minimal, (Vn)a tends to zero and (Vv)a can be seen to
 equal one. Thus for solvent-based inks, VOC = V * 5.388. Through a similar determination, the value of VOC
 = V * 1.31 can be found for water-based inks.

         Adjustments to these equations may be necessary as water-based inks from various manufacturers are
 used. Both equations will be necessary for emission monitoring In that it is the intent of Lustreprint to work off
 the existing inventory of solvent-based inks over time and the water-based inks will still contain minor amounts
 of volatile components.

 'Solid Wastes

         The operations at Lustreprint generate, in addition to emissions, a solid waste designated as F003, non-
 halogenated solvent. This waste stream is made up of unused inks, make-up solvent, and cleaning solvents.  The
 waste material is presently sent off-site for incineration.  The overall impact of water-based ink use on the solid
 waste stream generated on-site was monitored and compared to historical data for solid waste generation and
 disposal.

 TECHNICAL APPROACH

         In order to complete  the evaluation of the ink substitution on waste generation, it was necessary to
 monitor ink usage  and solid waste generation over a set period of time.  Although the original work plan
 proposed two one-week periods of data collection, four one-week periods were used to obtain the necessary data
 for the evaluation.  Historical data for emissions and waste generation was extrapolated for comparison with the
 weekly experimental data.

        The information necessary for project evaluation was recorded for each distinct printing job during the
 evaluation period.   Historical information was available for all operations from both printing presses.  The
 inventory recording and computerized billing systems presently used at Lustreprint allowed the obtaining of a
 list of the types and amount of ink and make-up solvent used for each printing job. Ink usage is monitored on
 per job basis for invoicing purposes.  Lustreprint provided the ink manufacturers' MSDS forms for information
 on  the  type and amount of VOCs present in the  ink used (these were  not directly analyzed and  the
manufacturers' information was relied upon). Information on the total volatile organic compounds used during
the test period was derived from the above.
                                                 15

-------
         To obtain a figure for the total emissions, through material accounting, information on the generation
 of waste liquid VOCs was required.  To accomplish this, Lustreprint segregated all waste liquids generated for
 the test period. The waste liquids consisted of waste ink, make-up solvent and cleaning solvent. The waste liquid
 was accumulated in clean, unused 55 gallon drums. To obtain an accurate figure for percent volatile? hi the solid
 waste, sample acquisition took place at the end of each test period as defined in the Quality Assurance Project
 Plan.

         The data collected from the ink usage and waste analysis allowed for a calculation of the VOCs which
 had been released as emissions  from the printing  process.  Although analytical measurement for emission
 generation from the printing process would be preferred to a material accounting calculation, it would have been
 beyond the scope of the present study to isolate the emissions from the different potential sources at the f aeility.
 The material accounting approach provides the most accurate information for evaluation of the substitution of
 the solvent-based ink based on data available and that which could reasonably be obtained.

        The remaining weeks of testing were completed in accordance with the work plan.  These allowed an
 evaluation of the emissions calculated versus the  amount  of solvent- and water-based  ink used.  It  was
 hypothesized that, on a per pound basis, the week utilizing the smallest amount of solvent-based ink would result
 in the least emissions.

        Further evaluation of the data was considered since other variables may have an impact on ink usage
 and therefore, total emissions. Total production at the facility was determined by pounds of finished product.
 The total production of printed material was recorded as part of this investigation in the hope that this would
 be  a useful parameter against which to relate emissions to. However, it was determined that an evaluation of
 emissions per pounds of product produced would be inappropriate due to the variability in the percent of web
 material surface which is covered by the printing process. This percent varies from over 100% coverage to a
 small fraction of the surface depending on the customer requirements. Additional complications arose from the
 overlay of inks, as is required for some jobs.  As a result, it was concluded that emissions generated versus the
 pounds of ink used would be the most representative basis for the present evaluation.

 TECHNICAL EVALUATION

        Due to the number of VOC sources located within the Lustreprint facility, analytical measurement of
VOC emissions from the Hudson/Sharp press would have been impractical.  Therefore, a derived material
balance calculation was selected as an alternative means to determine emissions from the Hudson/Sharp press
to evaluate the impact of the use of water-based inks.                                        :

                                                 16

-------
         A total of four (4) one-week long study periods were completed to acquire the information on ink use
 for the technical evaluation of the ink substitution.  Routinely, during the course of each week, several printing
 jobs were completed according to customer demand and work schedules.  Numerous aspects associated with each
 individual job may impact VOC emissions such as the percent of water-based ink used, percent coverage of the
 substrate material and quantity of material printed.  The one (1) week long tune period was chosen to collect
 data for evaluation from several jobs. Data for all printing jobs completed during the study period were collected
 and used for VOC calculations. This niinimized those aspects which would result in variance of VOC generation
 from individual jobs.

         The type and amount of ink used for each printing job is recorded by Lustreprint personnel on a Job
 Ink Use Work Sheet and a Daily Operations Report Form. These forms are completed by the press operator
 at the time of job completion and include both ink usage and make-up solvent added during the course of the
 print run. This information is  then processed by a  computer billing system to provide a total picture of the
 printing job with respect to material usage.

        These forms were obtained for the printing activity for the four weeks of the study.  For three of the
 weeks, both solvent- and water-based inks were used. The four weeks were: March 18r23, July 15-20, July 22-27
 (during which no water-based inks were used), and September 23-28, 1991. The information was transferred to
 the WRITE Ink Usage Report identifying ink stock number, ink type (water versus solvent and color), pounds
 of ink to press, pounds of make-up added, pounds returned to inventory and weight percent VOC hi the ink.

 VOC Calculations

        A material balance calculation was  used to determine VOC emissions resulting from  the printing
 process.  Figure 2 provides a Process Flow Diagram of flexographic printing. Sources of VOC for the process
 include the inks and make-up/cleaning solvent. Ultimately, VOCs are either retained hi their liquid form and
 returned to inventory as unused ink (RTI Ink), wasted as unusable ink (combined with cleaning solvent) or
 released as point  source or fugitive emissions such that:
        VOQ +  VOQ =
where,
        VOC,                  =      the weight of VOCs in the ink brought-to-press,
        VOCs                  =      the weight  of VOCs in  the  make-up solvent (including small
                                      quantities of fresh cleaning solvent),
        %VOCadj(Irt!)           =      the quantity of VOCs  in the ink returned to inventory,
        %VOCWMle(W)          =      the quantity of VOCs  in the waste ink,
                               =      the weight of VOC emitted.
        Trace amounts of VOC entrained in dried ink on the printed material was assumed to be negligible for
purposes of VOC emission calculations.

                                                17

-------
1
rt
a
.22
0)
«• o
•§

fl
^ 1
s
w
"5











"c"
o
••§
g J
1
g M\

L
CO
_0
CO
E
LU
O
0
4- >
"o "IT
g 8


/•-I ^
.2
CO



TJ
2
a.
o
s
. 1
L
1 	
I
I
I
I
^ 	 ,,
1 1
c
~ 1
g ,
8 '
X __,

^ 1

1
1
1
i i
CO
1 1
OJ 1
° *l '
I
I
I
I 	
^ Product
— -


i

i

f

c


[^
i
CO
o
cS
j
0
Q.
(D
Q.
C0_
L
1

^
3
»>»

i


CO
s
Q.
CL
J

*>^
O
1
L

— "1
1
1
1
^ 1
1
I
I
i "c

If" O
— CO
J . ^
- 1 |
1 I
- .
1

1
1
1
1
1
1
1
	 |
i
>



E o> |
<3 c 2
*-• *"• m
m r~ ^**
z* .— O
CO i_ u_
- 5 Q- cu
® >. O UJ
^ > 'F tl:
o <~» -1— rr
S £&i
r- fc.— ">
•§ to co >
« CO o i
0) CD 5^ ^
O O m O.
o . - rri
2 u_ ty
°- o >»
J_2 4— «
. CD C
CM > 13
£?<§
2. .^ fl\
iZ :> u]


cd




I
     18

-------
Ink Volatile Content-
        To calculate the input of VOC from the ink generated from the ink use, it was necessary to first
determine the weight percent of volatiles in the ink used.  Many of the inks used by Lustreprint are supplied by
the manufacturer in  a ready to use form. For these inks, the weight percent of the volatile components is
provided in the Material Safety Data Sheets (MSDSs). The types of solvent materials commonly used in solvent-
and water-based ink  formulations were found in the MSDS and in the RECRA Environmental Incorporated
analytical reports. For ready-to-use inks, the pounds of VOC input is simply the pounds of ink used multiplied
by the total weight percent of volatile organic compounds in the ink formulations. However, several of the inks
used were formulations mixed by Lustreprint in order to meet customer requirements for specific colors. In
these instances, a preliminary calculation was necessary to determine the percent volatiles of the ink formulation
brought to presses as follows:

        The percent  volatiles of the base color inks were obtained from the manufacturer's MSDS.  The color
formulations, hi weight percent, were provided in a printing industry standard book entitled PANTONE Color
Formula Guide5. Ink formulations typically consist of equal amounts of base ink and varnish.  The type of
varnish used for the formulation  is  a function  of the application  of the ink, Le., surface versus lamination ink.
The code, identifying the particular color formulation and varnish used, is listed on the Lustreprint Job Ink Use
Work Sheet as part of the ink identification number. The percent volatiles of the ink formulations brought to
                                               *
press was then calculated using a weighted average of the various components. (For sample ink calculations see
Appendix A.)

Percent Volatile Adjustment for Make-up Solvent

        Make-up solvent or thinner, consisting primarily of alcohol  and acetate solvent materials, is added to
the ink prior to  start-up of the printing process to obtain the proper ink viscosity. An air diaphragm pump
circulates the ink from the supply bucket to the ink pan until an even consistency is achieved. Additional make-
up is added as necessary during the press run.

        Make-up solvent added to  the inks represents another source of VOC in the printing operation.  An
adjustment to the initial percent volatiles of the ink brought to press was then required to account  for the
addition  of make-up  solvent during the printing process.

        The amount and type of make-up solvent used for each ink was documented on the Lustreprint Job Ink
Use Work Sheet and is provided in the WRITE Ink Usage Report.   Several types of make-up solvents are

                                                 19

-------
commonly used by Lustreprint, all of which are 100% VOC with the exceptions of water (0% VOC content) and
the material  identified as 8020.  This is a make-up material developed specifically for water-based inks and
contains 20% VOC.                                                                     !

        The  adjusted percent volatiles for each ink brought to press was calculated using the pounds of ink
brought to press, the pounds of make-up solvent added and their respective volatile content.

VOC Output Totals
                                                                                        I
        With the adjusted percent volatiles determined for each ink used during the study period, the total VOC
input to the system could be  determined. However, this figure did not represent the VOC emissions for the
study period. In addition to VOC released as point source or fugitive emissions, VOC from inks and make-up
solvent may either be returned to inventory as unused ink or disposed as waste ink.            ;

        To calculate the pounds of VOC returned to inventory as unused ink, complete mixing of the ink and
make-up solvent was assumed. The percent volatiles for the ink returned to inventory (RTI Ink) was therefore
calculated for the ink after adjustment for the addition of make-up solvent. With a value for the pounds of ink
returned to inventory reported on the Ink Usage Report, a value for the pounds of VOC in the ink returned to
inventory was determined.

VOC Content of Solid Waste

        Waste  ink and waste cleaning  solvent is another end point  for unevaporated VOC which must be
deducted from the total VOC input to determine the pounds of VOC emissions.

        In order to minimize  solid  waste  generation,  ripening solvents at Lustreprint  are reused wherever
possible.  Depending upon the cleanliness of the solvent residue after use, the material may be placed in one
of three locations:

         a)  Relatively clean  solvent  is placed into solvent ink returned to inventory to reduce the amount of
         make-up needed of the next time the ink is used. This represents the majority of cleaning solvent and
         is reported in the Lustreprint Job Ink Use Work Sheet as make-up. This cleaning solvent Was accounted
         for  in the material accounting by the adjustment of the initial percent volatiles for the ink.
                                                  20

-------
        b)  Cleaning solvent which is not suitable for addition to the ink, but not spent, is placed in a color
        dedicated solvent bucket for equipment cleaning, i.e., pump and hose cleaning.

        c)  Finally, solvent spent beyond its useful life  is  collected in 55 gallon drums and sent off-site for
        incineration as POOS waste.

        To account for unevaporated VOC, the generation of waste ink and solvent was monitored. Both water-
based wastes and solvent-based wastes were placed in separate clean 55 gallon drums, during the study period.
At the end of the week long study period, a sample of the wastes was obtained and analyzed for VOC content.

        Sample collection and analysis was completed by RECRA Environmental Incorporated, according to
the approved Quality Assurance Project Plan for the project. Proper storage and preservation protocols were
followed and the analysis was completed within the specified holding times for the sample material The samples
were collected using an open tube sampler (thief). Pre-cleaned glass 40 ml volatile vials with Teflon septa were
completely filled with zero headspace.

        The analysis utilized method 8015 (GC with flame ionization detector) for non-halogenated volatile
organi.cs (methanol,ethanol, propanol, isopropanol, l-methoxy-2-propanol and 2-(2-butoxyethoxy)ethanol) and
Method 8240 (GC/MS) for volatile organics  (ethyl acetate, n-propji acetate, isopropyl acetate  and  methyl
isobutyl ketone) per the USEPA "Test Methods for Evaluating Solid Waste - Physical/Chemical Methods" SW-
846, Third Edition, November 1986.  Duplicate analysis and method blanks were utilized to assess accuracy of
the analysis.

        Sample analysis for solid waste was only required for two of the four study periods. There was no solid
waste generated for Weeks 3 and 4. Activities for Week 1 only generated a small amount of solvent-based solid
waste  while Week 2 resulted in a small amount of water-based solid waste. Waste tracking for material
accounting calculations required that the press operator complete and initial a waste generation sheet, providing
information on the type and quantity of waste being disposed.                           ,

        The analytical results, along with the recorded amount of waste material, allowed for the calculation of
the VOC  contained in the waste in liquid form. This waste VOC figure was then subtracted from the total VOC
along -with VOC in the ink returned to inventory to calculate a final VOC emission amount using the equation:
                                                  21

-------
        OU« + s -

where, in addition to the parameters previously described
                                                    •                                     i
        I>Urt             =      the weight of the ink used brought-to-press

        S               =      the weight of the solvent used for make-up and cleaning

        !„,              =      the weight of the ink (including make-up solvent) returned to inventory

        W              =      the weight of the waste generated

        %VOC.g        =      the weight percentage of VOC as adjusted (This number was determined from
                               the VOC quantity in the MSDS for each ink used to make up a mixture,, The
                               %VOC«s  = quantity of VOC in the ink, plus any make-up solvent VOCs,
                               divided by the quantity of ink plus the quantity of make-up solvent.), and

        %VOCWMto      =      the weight percentage of VOC in waste (%VOCWMto was determined by
                               analysis of the VOC content in the waste collected.)


        As an example during week i, Job 34946, which generated no waste, had the following values:


        1^             =      216 pounds, of which only 1.2% was VOCs,

        S               =96 pounds, containing of 20% VOCs, and

        Irt,              =93 pounds.


        From these we determine %VOC.dj  =  ((216)(l-2) +  (%)(20))/(216 +  96)  =  7.0%, and further,

VOC^fcrf =  (216 + 96 - 93) * 0.07 = 153 pounds.


        Further calculations are detailed in Appendix A.


        These methods for tracking,  calculation and analysis completed the derived material  balance  to
determine VOC emissions for the study period. Minor losses due to spills, VOC in printed material, absorption

by cleaning rags, etc, were assumed to be negligible.  These methods were repeated for each of the four weeks

of study. The information was used to graph VOC emissions versus pounds of ink used.  This data, along with

historical data provided by Lustreprint, was  used to calculate a theoretical value for  VOC emissions for

comparison which provided insight to the impact  of water-based inks on  total VOC emissions at the facility.
                                                  22

-------
                                           SECTION 5

                                            RESULTS

HISTORICAL DATA

       Lustreprint is required to submit to the New York State Department of Environmental Conservation
(NYSDEC), a monthly report describing VOC emissions from the plant as a result of operations.  This
information includes the amount of ink used with the necessary calculations to determine total VOC for the
month and is used by the NYSDEC to determine regulatory compliance.

       Historical data was acquired for the period from April through August of 1990. This period was chosen
for comparison with  WRITE data for  several reasons.  A three shift work schedule had been adopted by
Lustreprint at this time providing comparable ink usage. The corona treater and other equipment had not yet
been installed  for the changeover to water-based  inks.  Therefore, the information represents a period of
exclusive solvent ink use. Lastly, the data represents a period with no significant plant operation upsets which
may impact ink and solvent use. The historical data, in its raw form, represents total VOCs from ink and make-
up solvent use for each month from the two presses used by Lustreprint. It was estimated that half the ink usage
could be attributed to each press.  Therefore, the data was adjusted to represent VOC emissions from a single
press (total VOC/2)  for a one week period (total monthly VOC per  press/4) for comparison with the data
acquired during the study period.

       The data hi Table 1 shows the VOC emissions, as a function of ink use, based on historical data.
                                 TABLE 1.  HISTORICAL DATA
                                              1990
                                             Ibs. of ink             Ibs. of VOCs
                                             used                  calculated
                      Month                 Per Week             Per Week
                      April                 3,038                 2,111
                      May                  1,681                 1,700*
                      June                 2,686                 2,289
                      July                   2,109                 1,731
                      August                 2,945                 2,345
       *       Although this is a practical impossibility, due to this being a calculated value derived from the
               historical operational data recorded, the calculated value exceeds the amount of ink used.
       This data is arranged in graphic form in Figure 3.

                                                23

-------
                                   SO
                                   O
                                   c
                                   .2
                                   "55
                                   at

                                   HI
                                   b
                                   O
                                   TJ
                                    0
                                   m
                                    C

                                   i
                                             CO
                                             «
                                             =1
                                             _w
                                             U.
24

-------
OPERATING DATA                   f  ^         . #


        Table 2 provides information on the number of inks used, total pounds of ink used, pounds of solvent-

based ink, pounds of water-based ink, percentage of each, calculated VOC emissions and VOC emissions as a

percentage of ink used, for each of the four (4) one-week long periods of the evaluation.

                                      TABLE 2. WRITE DATA
        No. of
        Solvent Ink (Ibs)
        Water Ink (Ibs)
        Total Ink (Ibs)
        Solvent Ink (%)
        Water Ink (%)
        VOC Emissions (Ibs)
        (calculated)
        VOC Emissions (%)
        (% of Ink Total)
        Waste (Ibs)
        Waste VOC Content (Ibs)
WEEK1
(3/18-3/23)
23
1,112
1,251
2^63
47.1
52.9
827.5
WEEK 2
(7/15-7/20)
32
1,746
508
2,254
77.5
22.5
1,251.7
WEEK 3
(7/22-7/27)
33
2^52
0
2,252
100
0
1^571.5
WEEK 4
(9/23-9/28)
22
549
688
1,237
; 44.4
55.6
509.0
35.0

55.6
543
55.5

20.0
 4.7
69.8

 0.0
 0.0
41.1

 0.0
 0.0
        A comparison of ink usage by type is shown in Figure 4


ANALYTICAL RESULTS ON SOLID WASTES


        Solid waste is generated as a result of unusable ink and spent cleaning solvents.  The operator who

generated the waste was required to identify the job, date, source, weight and cause of the waste and initial the

sheet. The information gathered for the evaluation of the ink substitution at Lustreprint did not provide a direct

correlation between the use of water-based inks and a reduced generation of hazardous solid .wastes. Solid waste

was generated for  only two  of the four weeks of study.  Printing operations during Week 1 resulted in the

generation of 55.6 pounds of solvent ink waste while Week 2 generated 20.0 pounds of water-based ink waste.

To improve waste tracking after the first week of the study, a Waste Generation Form was placed at the point

of disposal for solid wastes.


        In addition to the analysis of printing ink wastes, an analysis was performed on cleaning solvent used

for cleaning out the ink pans during week 4.
                                                 25

-------
                                   JQ

                                   03
                                   O
                                    c
                                   LU

                                   O
                                   O
                                   S
                                             «*•
                                             U.
26

-------
Week 1-Solvent-based Ink Waste        .-,   , .           ^
                                       *   fs-          f

        Although water-based inks were used during week 1, there was no water-based ink waste. Analysis was
carried out for the following solvents:                                                ,
                               TABLES. INK WASTE COMPOSITION
                       Component                         Weight Percentage
                       Methanol                                          0.8
                       Ethanol                ,                           1.0
                       Propanol                                         57.8
                       Isopropanol                                       20.0
                       2-(2-butoxyethoxy)ethanol                           0.4
                       l-methoxy-2-propanol                               1.8
                       Ethyl acetate                                       03
                       n-Propyl acetate                                   15.5
                       Isopropji acetate                          Not detected
                       Methyl isobutyl ketone                              02
                       Total                                             97.8
                       The balance is assumed to be solids                  2.2
        Information on solvent constituents hi inks was provided by Lustreprint at the start of the investigation
and did not include the following relatively minor constituents.  As such, the following VOCs are listed in the
MSDSs, but no attempt was made to analyze them, since they were not part of the original Quality Assurance
Project Plan (many MSDSs were provided after work was initiated):  n-Heptane, Toluene, Propylene glycol
monomethyl ether, 2-(2-ethoxyethoxy)ethanol, VM&P naphtha.

        With limited  use and quantity of these  VOCs (usually less than 5% of the VOCs in the ink), it is
assumed that their participation is very limited and thus would not significantly affect the results.

        For the analysis of the waste from week 1, two samples were taken from the waste drum as some settling
had occurred. The waste was mixed to the best  extent possible, then the samples were taken.  The results of
the samples were averaged to reduce differences due to a lack of a homogeneous mixture.
                                                  27

-------
Week 2-Water-based Ink Waste

        Although solvent-based inks were used during week 2, there was no solvent-based ink waste.

        The following solvents were determined to be present in the water-based ink waste:

Methanol, ethanol, propanol, isopropanol, 2-(2-butoxyethoxy)ethanol, l-methoxy-2-propanol, ethyl acetate, n-
Propyl acetate, and isopropyl acetate. VOC content was calculated at 27.7%.

        Methyl isobutyl ketone was not detected.  Again the remaining five VOCs were noi analyzed.

Week 3-Cleaning Solvent

        For reference, the solvent used for cleaning was analyzed. The following VOCs were determined to be
present:

Methanol, ethanol, propanol, isopropanol, l-methoxy-2-propanol and ethyl acetate.             ;
                                                   t
PRODUCT PERFORMANCE                      :

Product Quality

        During Weeks 2 and 3 of the study, information on the Quality Assurance checks employed by
Lustreprint was collected. The information consists of operator sign-off sheets for each printing job. The press
operator is required to check the printing quality for several parameters at the start and during the press run.
This check may occur several times during the printing process depending upon the size of the job.

        Ink color, print position and register are checked to assure a satisfactory product that meets customer
criteria.  Tests of printed material to check ink adhesion include the industry standard 610 Tape Test. A Tear
Sheet sample is collected for each QA check. The print is also checked for various printing imperfections such
as phi-holes, halo, fish eyes,  and roll marks.

        The press operator initials each quality assurance check as completed.  Problems, when detected, are
noted on the sheets.  A review  of the quality assurance sheets indicates that the use of water-based inks did not

                                                                                        i
                                                 28

-------
change product quality. Test runs previously completed by Lustreprint, using various inks on different plastic
substrates, likely precluded serious quality assurance problems prior to implementation.

Ink Durability

        Ink durability is defined as its ability "to withstand the condition to which it will be subjected on the
proposed  application".*  The use of water-based inks commonly results  in  the degradation of ink durability
characteristics due to inadequate adhesion. Several different tests may be employed to assess the durability of
the water-based inks. However, only one, the pressure sensitive cellophane tape tes{ was used as a gauge of ink
adhesion in  this study as it is the only one commonly used by Lustreprint. (See Appendix D for a description
of this test.)

Ink Appearance

        In some cases, water inks were successfully used, providing satisfactory printing results, with problems
arising after use by the customer.  Depending upon the ultimate use of the packaging, what  the package
contained and the means by which the packages were sealed, some difficulties were noted. Heat and stress of
the printed package material caused by the package folding and sealing process  at times resulted hi a loss of ink
adhesion.

        Some combinations of water and solvent inks were found to be incompatible. Water inks did not provide
a consistent opaque white for laminations to cover metallized films and resulted in "blocking" (or transfer of
print) when  printing on saran coated materials, especially cellophane.  In most  cases however, depending upon
the surface printed, there was no difference noted with the use of water-based inks.

        The addition of normal propyl alcohol in small amounts (less than 1%) to prevent ink foaming at the
ink pan and to assist in ink wetting was found to be beneficiaL Variations of the pressure sensitive 'stickyback*
material used to attach the printing plates to the plate cylinder, solid versus cushioned sticky back, also enhanced
the printing of solid plate backgrounds without pinholing.  The plate material may also have an impact. Photo
polymer plates work  well with  water but are more expensive than  rubber.  Nylon plates  are  a possible
compromise with a longer life than rubber plates.
                                                  29

-------
                                            SECTION 6
                                                 «

                          DATA SUMMARY AND ANALYSIS (DISCUSSION)

HISTORICAL DATA

       The historical data is graphically depicted as pounds of VOC as a function of pounds of ink used in
Figure 5.

OPERATING DATA RESULTS & DISCUSSION

       A discussion of the impact of water-based ink use on VOC emissions, along with solid waste generation
follows:

VOC Emissions
                                                                                       I
       The WRITE study data is plotted as pounds of VOC as a function of pounds of ink used and th'$ is
superimposed upon the historical data in Figure 5.                                         \         •

       A linear regression was completed to obtain a .best fit line for the historical and solvent only WRITE
data.  The linear regression resulted in an equation of the following form:

       y » 0.725x + 133,
       with x = pounds of ink used and y = pounds of VOC emissions.

       This equation will apply only to this study and is not a general equation to be applied in all cases with
all inks used at various facilities.

       According to this equation, 0.725 pounds of VQC emissions was found for each pound of solvent-based
ink used. This equation fits well with the adjusted percent volatiles calculated for the solvent-based inks which
generally ranged between 70 and 80%. The correlation coefficient for a linear regression is a statistic that
indicates how much of a change in the  dependent variable  (VOC Emissions)  results  from changing the
independent variable (Pounds of Ink). The closer the value is to 1, the closer and more predictable the
relationship between the variables. The value given for this parameter in the regression is 0.896, showing some,
but not strong, correlation.                         ;                                     ;

                                                 30

-------
                                            APPENDIX D
                               Pressure Sensitive Cellophane Tape Test

        This test is a comparison of the strength of the bond of the ink to the tape and the ink to the web
material. One inch wide 3M No. 610 High Tack cellophane tape is used. A six inch length of tape is applied
over both single and multitude colored areas of the printed surface. The tape is then removed at both slow and
fast speeds for a qualitative appraisal of the adhesion quality of the printed material.
                                                 68

-------
APPENDIX D
    67

-------
                                            APPENDIX C

        The Critical Surface Tension (CST) is the pertinent characteristic of the film material for ink application.
The concept  of surface tension for a solid is  discussed in an  article entitled "Surface Energies of Printing
Substrates" by Edward Charogoff, (FLEXOgraphic Technical Journal, October 1982). Charogoff explains that
surface  tensions for solids require an indirect measurement of surface energy.  The characteristic can be best
described in a simple experiment.  If one were to place a drop of water on a newly waxed automobile, the water
would bead.  A similar drop of windship washer fluid, due to its composition, would produce a larger area of
wetting, reducing the contact angle produced by the surface and the leading edge of the fluid. A drop of gasoline
on  the same surface would wet until the contact angle would be nearly undetectable. The Critical Surface
Tension of a solid film is defined as the surface tension of a liquid which gives a zero degree contact angle with
the solid film. By varying the surface material used in the experiment, a method to determine the CST of a solid
is possible.  The formal method devised for measuring  Critical Surface Tension (CST) is known as "Wetting
Tension Test", ASTMD2578.
                                                 66

-------
APPENDIX C
    65

-------
Savings from Variable Cost

       $24,587 X.60  =  $14,752 after tax

       $14,752X6.14  =  $90,577 NPV

Operation and Maintenance

       Assume $1,000 per year

       $1,000X6.14  =  $6.145 NPV

Project Net Present Value of Savings 

       Equipment Purchase           $< 62,901 >
       Operation and Maintenance      < 6,145 >
       Tax Savings on Depreciation      17,634
       Savings on Ink and Solvent        90.577

       Project Net Present Value    =   $39.165
                                                64

-------
      PRESENT VALUE CALCULATION
 Assumptions:
Equipment Life
Scrap Value
Tax Rate
Discount Factor
10 years
$0
40%
10%
 EQUIPMENT DEPRECIATION CALCULATIONS
                                      Straight Line Depreciation
        Depreciation/Year
               6,290
                      Tax Savings
                        2516
                      NPV Factor
                         6.14
                      Tax Savings
                                 Sum of the Years Digit Deprec
        Year
         1
         2
         3
         4
         5
         6
         7
         8
         9
         10
       Depreciation
          11,437
          10,293
           9,149
           8,006
           6,862
           5,718
           4,575
           3,431
           2,287
           1,143
Double Declining Balance
(Rate - 20%)
(Straight Line After 6 Years)
        Year
         1
         2
         3
         4
         5
         6
         7
         8
         9
        10
       Depreciation
          12580
          10,064
          8,051
          6,441
          5,153
          4,122
          4,122
          4,122
          4,122
          4,122
       Tax Savings
         4,575
         4,117
         3,659
         3,202
         2,745
         2,287
         1,830
         1,372
          915
          457
     Factor
      .909
      .826
      .751
      .683
      .621
      .564
      .467
      .424
      386
NPV
4,159
3,401
2,748
2,187
1,705
1,290
 939
 641
 388
_176
                                                                TAX SAVINGS =   17,634
       Tax Savings
         5,032
         4,026
         3,220
         2,576
         2,061
         1,649
         1,649
         1,649
         1,649
         1,649
     Factor         NPV
      .909          4,57"4
      .826          3,325
      .751          2,418
      .683          1,759
      .621          1,280
      564           930
      513           846
      .467           770
      .424           699
      386           637

TAX SAVINGS =  17,238
Use Sum of the Years Digits $17,634
                                                63!

-------
                                            APPENDIX B

ECONOMIC CALCULATIONS

Fixed Costs $62,901

Savings from variable costs

From LustreprLnt Inventory for 1st - 3rd quarter, fiscal year 1991-1992.

                                      Use (Pounds)                          Avg. Cost/Pound
        Solvent Ink                      56,839                       $2.78
        Water Ink                       15,030                       $2.15
        Make Up Solvent                72,517                       $0.57
        Water Make Up                 3,160                       $0.10

Savings on Ink

        15,030 X $0.63/pound =   $9,468.90

Savings on Make Up

        72517 # Make Up/56839 # Solvent Ink = L27 #

        1.27 # X 15030 # Water X $.47  =  $8,971.41

Total Savings (Aanualized)

        ($9,468.90 + $8,971.41)/0.75  =  $24,587.08

PAYBACK CALCULATIONS

Fixed Cost

Variable Cost Savings  =  $24,587  =  2.56 years

Disposal Costs
        From Lustreprint cost per 55 gallon drum
               Disposal       $105
               Transportation  $ 45
          Total Disposal Cost  $150/drum

Liquid Waste Disposal - 18 month average (April 1990 - September 1991)

        Average  =  314.7 gallons  =  5.72 drums X $150  =  $858/month
                                                      =  $10,296/year

Estimated Ink Splitter Cost - $8,000

PAYBACK CALCULATION 2

Assumes purchase of ink splitter and virtual elimination of hazardous liquid wastestream.

        Fixed Cost               62.901 + 8.000
        Variable Cost Savings  =  24,587 + 10,296  = 2.03 years
                                                62

-------
APPENDIX B
     61

-------
«.  ,«ttn<-
-------
                                                          »-*  *.   O
                                                                                                     o   a    *   «
                                                                    00        «

                                                                    *   2        S
                                                          *-*  *n   o
                                                     ^-   <-*  O   O
                                                                                                    *-   ^x  o   «   <-i   o   o   «n
                                                                                                    £**>WW40W«>
s
5
                                                               g   g   g
                                          g   g   g        g   g   g
                                                     3   *"».   
                                                     «   .a  M   ^

                                                                   5   S
                                                                                                             o   to        to
                                                                                                                            
-------
5
                                                                    *»   ^  01   O
                                                                          t«
                                                                         .a
I
               g        g        g


               g        §        §
                                                                    A   5
I
               g   g   g   g   g   g
                                                g   g
 3   ,—s       U>   »•»   «*
                              s   s
               s   s
                     '
                                                                    **   O-H  «\   «r
                                                                     L.    *   ifi   tD
3   2

X   in
                                                                                    s
                                                                                    u.
                                                                                    X
                                                                                        58

-------
      *•*  *si   «r    o
                                                                             -^   o   *•>    o
                                                                       "P   *•"»   m    •»
                                                                        *    •   CM    «
r.        S   IS    3S   S   g    fi
                                                                         g    g   g
g    g   g   g   g   g
§    g
                                                                      g         g   g    g   g   g
                                                                                                                             g         g   g   g    g    g   g
                                                                     J!   *
                                                                                                                                      3   A   rs.   *•*   <*)    «*>    mo

                                                                                                                                      •    •vwnniSJS
               g   8
                                                                                           S    5
                                                                                                                                                                g   g
•^   o    r^
f*.   Csl    -t   ^
«•   -j    o   o
P   M    CO   W
o   X    o.   -I



 •   I    ;   !


;*    >~    M   CD
                                                                                                                                              r*-    m   o    «
                                                                                                                                              «        
-------
                                                         t   5
                                                                                             •o   <^
                                                                                              •   *
                                                                                              ;   5
^       S   S   S   15
 •        CJ   W   O   «*
o        «   r*.   o   o
I
                    g   g   g
                                                         1   I
                                                              •
          g   g   g   g   g
 8        «
Z        8
                                             g   g
                                                                                                                                       |
g
§
£   -O

  *   &
                                                                                             b    «   
-------
      £,  «   "0    »   »
                                                                                                  o    JB   •-•   *.   »4   m
 •e  *-*   *o   o   «n   «>
 •   *   »•*   f*   o»   to
 M  ••
 a  t-
                                            ||   -   3
                *   *   «
I
                                                                s
                                                                                                           g   2   S   g
               g.   g   g
               i   s   g
          g   g
&        §   §

                                                                    •
                                                               g   g
                                                                                                11
                                           u        g   g   g   g

                                           K        S   8   §   g
                                                          ^N   IA

                                                          JS
                                                                                                 3   *-»  fti   •*•
              «•»        in
                                                                                                «        S   S   g   g
    *~+   *o   o    «o
                                                                                                t    «
Whit* 10U471
R«d 33RFS161
o
o
, s ""•
3 \ ;. -
> «a
1 1
*•« Ot
• Jt
1 1
                                                                                                         1   i   5   i
                                                                                                         «   «n   ir
                                                                                      55

-------
                                o  A  a>  *•»  to
                                                              ^"»  «  »<  X
      r>  f^.  »  <
|


^
                                        g  g

                                        §  §
                                                                g g  g
      g  g  g  g
                               JS  I
                                     g  g g
                                                           ii
                                                                g g  g  g
                                                           ai ^> N

                                                           E S "
         J*  X 5«
         »  S V
                                                  54

-------
      «7*   CM   10
                            f)

                            *-t
                            CJ
                                                                                                                    O    A   M   O   «   *-<   »•*
                                                                                                                   *•    *-   CM         •-«   O   «-•
                                                                                                                         *—•              M   •-*   •«
                                                     *-*  CM   m   •-«    o
                                                     S   2   *   ~    *
                                                                                                             *-»  o   co
                       g   g
                       V   m
                       r»   os
                                               -        s    g   g   a;
I
                       £   £
                       o   o
                       o   o
                       o   o
                                                               g   g   g
 3   «—«   -•    CO
                                                3   *••*         *>   C»   «
           g   g
                                                                      g   g   g    g
g   g   g   g   g
                                                U         rfJ
                                                               g    g   S
                                                               ID    ID   ID
                                                                                                                  ^   in   in   m
                                                                                                        ««   O-K   *«•    CM   «3   ^.   •-*
                                                                                                        U   M   CM    O   O   «   CD
                                                                                                        •   A   «    M   CM
X
o
trt
u_
cz

-------
                                       O*   O         -V
                                                                                      —   CM    O   «
                                                                                                                                         *-%   *n   o    o
•v*-*.ootr4O9eei
 •    *tors»«or*.*n*n
 «   A   -   >-
           c    c    c    c
          O   O   tfl   tv
          *r   «*•   r-*   «*
          m   m   r>)   «*>

          -*   _i   t/»   w
          o   o   z   x
          •-«   ^^   a.   a.
                                                                                     ^»   CM   «n   ^
                                                                                     c    c    c    c
§   s    s
*H   01    Ot
O*   CM    <7i
CM   •*<    CM


>   *    »
                                                                                                                                              •^   CM   r>
                                                                                                                                                                                                    \


                                                                                                                                                                                                    i
                                                                                               52

-------
           ot   m   «>
                                                   t   2
                                                                                                                                     ^   *-»  o   •-"

                                                                                                                                     t   £   *   «
           g   g   g
                                                  ?

                                                  i

                                                                                                                                     3

                                                                                                                                     i
                g   g
                §   §
 3   —•        00
                                                                                                                                     k   3
 g        s   g   s.
 >        o   o   o
X        CM   O   O
                                                                                                   g   g


                                                                                                   8   i

                                                  3^0,


                                                  E   JQ
                                                                                                                                     5   5
                                                                                                                                     
                                                                                                                                               •-<   r«.   »
                                                                                                                                               0%   m   n
                                                                                                                                                    &   &
                                                                                  51

-------
    30O  <*%  «O
    fS  —  O4

L«^  O\  f^•  *t
                                                                                         SCM  «*l
                                                                                         o*  —
S
                                                            *
                                       s  a  2  a
                                       s  *  s  ?
                                                       a a a
                                                       s ? ?
                                                                                                                                             a a
                                                        *O  09  f^  ^^
                                                        O\  &  rt  Ct


                                                        O  O  O  O
                                                        t»i  oo  O
                                                        "^  °*  S

                                                        odd
                                                                                                                000
                                       ^ c — in
                                       odd
                              O  O  Q  O
                              no  «n  O  <5
—  \o  m  o o
O^  f*^  t*-  O O
                                                                    2  o
                                         8  8
                                         2  8
                                                                                  8  S  8  8
                                                                                  r^  c
                                                     07

                                                     a,
                                                     e

                                                     O
           EL



           O
                                                                              a.

                                                                              "8
                                                                              06
                                                                                                 50

-------

y*


1
1
yt
1

1
5
jj«
|

5
j>

£?
o
CO
8
*~*
1
u.
)^
JO
0


ys
s
f^T
^
3
?

S
o

3
f~

s
ys ys ys
O ~~ v^
S t  r- •*
f^ C*»
gg
s s

M Q
0 0

*O O f,
•V O 00
CN 0 OK

8 2
ts ye t^ w? ye y? b^ t£ y? y? y? ye y? y? y? y^ ye y?
3O — "O ^tr^^v^ QOcJf^v) oor^o»/>*O

c^csr^ eN'r^r-4 c^c^cJc^ cs*C"4 C>
t^ tsi 0 0; vo ;« 0 CT> O O t~-' O O; M — •* O »

gg§ ^ §S§§
(S
1
u
o
*^H
tu
a.
2
tf>



1
j
s *
"< "8
z a:
Z"o
r*\
4
I


JS
m
CO
e
O
"o
g ^


OO
s
J :
K '
E
O
"5
3 3
f- H
o o
H H



1 1
g E
H m
"S "o
TOTAL







                                                49

-------
Some inks are used 'as is' and VOC content is obtained directly
                                         #  "
  from the MSDS. Other inks are composites and are calculated below.
  Lustreprint provided data for those with no MSDS
Number
Description
%VOC content
NONE
M
HSL
P-130130
201
302
304
PMS 355
Process 401
P-97735
W-129936
P136CW-130164
W-129100
W-130163
W-130140
W-130299
TB82-0666
403
306
61-V-29
61-V-3B
90U496
10U47I
7S4F
10L1540
50U48S4
PMS 476
Black (no MSDS)
Varnish(Surface Printing)
Varnish(Laminating)
White
Yellow
Warm Red
Red
Green
Blue
Yellow
Orange
Yellow
White
Red
White
Yellow
Purple
Blue
Rhodamine
Unknown
White
Black
White
Silver
Whita
Unknown
Brown
       11.30%
       56.53%
       70.00%
       50.00%
       75.95%
       67.35%
       73.90%
       73.40%
       74.25%
       70.30%
       11.30%
        5.30%
        1.20%
        5.30%
        7.30%
        5.30%
       69.50%
       74.30%
       69.60%
       56.50%
       70.00%
       55.00%
       42.50%
       63.80%
       44.00%
       56.60%
       73.50%
                                               48

-------
The following equation is used in the calculation of the VOC emitted and the quantity of inks used:

                                =       (IlUrt + S - IJ%VOC-j -
                IBM            =       Weight of ink at start

                S              =       Weight of solvent added as make-up

                !«;             =       Weight of ink returned to inventory
                               	       /\fg~\/~*  i_ \ff\j~* \ //T   i  O\                         '
                               =       (VOC,  + VOCs)/(IrtMt + S)

                VOC,          =       %VOC^IJUrt), where %VOC, = %VOC content of the ink

                VOCs          =       %VOCs(S), where %VOCs = %VOC content of the solvent

                W             =       Weight of waste generated

                        „.      =       Analytically measured VOC content of the waste
Note:   %VOC^- is the weight percentage of VOC as adjusted (This number was determined from the VOC
        quantity (weight percent) in the MSDS for each ink used to make up a mixture).

        VOCs is the weight of VOCs in the make-up ink including fresh cleaning solvent.
                                                 47

-------
APPENDIX A

-------
                                       IX.  BIBLIOGRAPHY

E. Charogoff, Surface Energies of Printing Substrates: Sun Chemical Corporation, FLEXOgraphic Technical
Journal, October 1982                             >

M. Capristo, Water Inks Can Offer Compliance. Cost Efficiency.  Performance Quality. BASF Corporation,
FLEXOgraphic Technical Journal, February 1986

R.A. Zuck. Aqueous Technology Can Work for Converting Uses: Paper, Film & Foil Converter, November 1990

R.M. Podnajny, Water-based Inks and Housekeeping. Converting Magazine, December 1989

J.V. Marra, W.E. Philhower. Waterbornc Inks Offer Advantages on OPP Film: Paper, Film & Foil Converter,
October 1987                                     |

R.W. Bassemir, R.  Krishnan, Practical Applications of Surface Energy Measurement in Flexogiraphy;
FLEXOgraphic Technical Journal, July 1990.

F.L. Jones. How We Anticipated and Corrected Problems When Converting to Water. FLEXOgraphic Technical
Journal, July 1990

J J. Morris, Low Solvent Ink Technology:  Where is it Now?  Where is it Headed?: FLEXOgraphic Technical
Journal, April 1986

A.M. Hadam. Water-based Ink - The Long Shot Pays Off; FLEXOgraphic Technical Journal, September 1986

J.V. Marra, W.E. Philhower: Water Inks on OPP Film: Effects of Process Conditions on Performance: Hercules
Incorporated Research Center; Wilmington,  DC, 19894
                                               45

-------
                                        VIII.  REFERENCES
                                       '  *fr  -      ".: ,f  '"'  -

1.       Guides to Pollution Prevention - The Commercial Printing Industry. U.S. EPA, Cincinnati, OH.  EPA
        Document EPA/625/7-90/008, August  1990

2.       D. Markgraf, Corona Treatment. An Overview. Enercon Industries Corporation; Menomonee Falls, WI
        53051                                                                     ;

3.       G.E. Ealer, S.B. Samuels, W.C Harris, Characterization of Surface-Treated Polyethylene for Water-
        based Ink Printability. Tappi Journal, Vol. 73, No. 1, January 1990

4.       G A. Makrauer, Innovations in Flexographic Printing; Reducing VOC's with Water-based Inks When
        Printing on High-Slip Polyethylene Films: AMKO Plastics Inc., Cincinnati, OH 1987

5.       PANTONE Color Guide

6.       Flexography:  Principles and Practice. 4th edition. Flexographic Technical Association, Ronkonkoma,
        NY
                                                 44

-------
        The economic evaluation for the project indicates that the decision by Lustreprint to substitute the 'water
inks for solvent inks was financially a beneficial one. To underscore the selection, a brief discussion of the
alternative which was available to Lustreprint is necessary.  The alternative to the substitution would have been
the installation of an incineration unit to control plant emissions.

        The estimated cost of a facility wide incineration unit varied significantly. The reported range provided
by Lustreprint was between $200,000 and $1,000,000. In addition, the VOC content of the Lustreprint emissions
would have been insufficient for proper operation of the incineration unit. It was estimated that the cost of a
supplemental natural gas feed for the unit would cost approximately $45,000 per year. Furthermore, this control
technology in lieu of waste minimization, would not have accounted for the potential for further regulatory
restrictions in solvent use for the printing industry.
                                                    43

-------
          The Sum of the Years Digits yields the most favorable net present value and will be used for the
  economic evaluation.


          The net present value for the cost savings assumes no adjustment for inflation, and no increased use of
  water-based inks resulting b a steady savings of $24,587 over the ten year project life.  Discounted 40% for tax
  purposes, the savings become a ten year annual cash savings of $14,752 for a net present value of $90,650.


          The cost of operation and maintenance of the equipment, including the corona treater, is reported to
  be nominal However, to account for utility cost and repairs, a cost of $1,000 per year is assumed.  The net
  present value of this cost is $6,145.


         Further adjustments to the project net present value could be realized  from savings accrued from
  improved waste handling.  However, since this has yet to be implemented,the adjustment will not be included


         The final calculation for the economic evaluation is the summation of the net present value for the cash
 flows associated for the project The cash flows and value are summarized below.         !
                                  TABLES.  NET PRESENT VALUE
 Depredation   Initial
 Method        Investment

 Straight Line   $62,901

 Sum of Years   $62,901
 Digits
Operation &
Maintenance
$6,145
$6,145
Tax Savings on
Depreciation
$15,461
$17,634
Savings on Ink
and Solvent
$90,577
$90,577
NPV
$36,992
$39,165
 Double
 Declining
 Balance
$62,901
$6,145
$17,238
                                                     $90,577
                                                      $38,769
Assumptions:    10-year life span
                no salvage value
                10% discount factor
                40% tax rate
                $1,000 per year O&M cost


        The positive net present value for the project indicates that the project will have a favorable impact on
cash flows and will ultimately result in a cost savings.
                                                42

-------
         The payback period could be further reduced by eliminating the solid waste disposal. With the complete
 changeover to water inks and the planned purchase of an ink splitter at approximately $8,000, an additional
 savings for solid waste disposal is possible. The payback period would then be reduced as noted in the table.
                                    TABLE?.  PAYBACK PERIOD
                Initial Investment       Projected Savings               Payback Period
                        $62,901                $24,587                    2J6
Current
Process
Revisions
Adding
Ink
Splitter
Full Water-
based Ink
Conversion
                        $70,901
                        $62,901
 $34,887
$117,078
2.03
0.54
        Reduced material handling, regulatory and training costs would lower this payback period further. It
was not possible to quantify these during the study and it is estimated that their effect would be minimal unless
full conversion took place.

Net Present Value-
        A more thorough economic evaluation for the project can be completed by calculating the Net Present
Value (Appendix B).  The project net present value is a summation of the cash flows for the project during the
projected life span taking into account the time value of money. A positive net present value indicates favorable
economics for the project A negative net present value indicates that the project does not meet the  criteria
selected for the project profitability.

        For the purposes of the calculation, it is assumed that the new equipment and press modifications,
including the corona treater, have a ten year life span with no salvage value.  A 10% discount factor  for the
internal rate of return will be used along with an assumed tax rate of 40%.
                                                    :                                      I
        The initial investment by Lustreprint for the corona treater, ancillary equipment and press modifications
is  $62,901.  The investment provides tax savings from equipment  depreciation write-offs which must be
determined The net present value of the tax savings is dependent on the depreciation method used. The three
most common methods used for calculating equipment depreciation are, straight line, sum of the years digits and
double declining balance depredation.  The net present value of the tax savings using the three methods are:
                                                  41

-------
Miscellaneous Savings and Costs-
        At the time of the development of the Work Plan, the changeover to water-based inks was expected to
have an impact on labor and overhead  expenses.  Savings and/or costs  were anticipated from insurance,
regulatory compliance and training cost.  However, the continued utilization and storage of solvent inks,  as
described earlier, complicated the task of delineating the cost variance for these parameters.

        Lustreprint did not report any increase in cost for labor or training as a result of water ink use. Training
for water ink  use was often supplied by manufacturer sales representatives.  Meetings were held with press
operators monthly, on a shift basis, to discuss any technical difficulties encountered on the shop floor.

        Cost reductions expected from insurance due to the deletion of solvent inks from inventory are unknown
and not anticipated for some time.

        Cost savings from regulatory compliance are reflected in the anticipated reduction hi permit fees  as
stated in previous section. Additional savings may be incurred as a result of the elimination of the preparation
and submittal  of monthly emission reports and periodic inspections by State environmental officials.

Economic Calculations

        The calculations for the economic evaluation of the Lustreprint project assumes a constant rate  of
production and ink  use.

Payback Period—                                                                    ,
        The simplest form of economic evaluation for a project results from a project payback calculation. This
calculation divides the fixed initial costs of the project by the projected annual savings. The result provides the
number of years it will take to recoup the cost of the initial investment (This is based only on the 21% water-
based ink usage level, although this rate could increase with the existing equipment.)

        For Lustreprint, the initial investment is the fixed costs from press modifications and equipment detailed
earlier in the report. Using the figures generated for cost savings at the 21% water-based ink utilization level,
savings anticipated for the year from variable costs include ink and solvent savings. The investment, savings and
payback period are summarized below in Table 7. (Permit costs have not been reduced for this calculation, since
this is based on 21% water-based ink usage. Until further reductions are achieved, the VOC levels may still
exceed the lower permit levels.)
                                                 40

-------
 Disposal Costs--
         The printing process generates a solid waste which consists of waste ink and cleaning solvent.  This
 material is disposed of by off-site incineration.  The most recent cost reported by Lustreprint for disposal is $105
 per 55 gallon drum and $45 per drum for transportation for a total of $150 per drum.

         At present, both wastes from water-based inks land solvent-based inks are disposed of in this fashion.
 Although  water-based  ink  waste (without any solvent  content) might be nonhazardous (depending on the
 presence or absence of regulated compounds such as metal compounds), Lustreprint does not release waste ink
 into the municipal sewer system due to a combined storm/sanitary sewer system in the area.  Once a complete
 changeover to water inks occurs at Lustreprint (followed by elimination of the solvent content of the water-based
 inks), an ink splitter could be purchased for approximately $8,000 to eliminate the cost of solid hazardous waste
 disposal                                                                                i

         From  historical data, over an eighteen  month period, the  average monthly waste generation for
 Lustreprint is 315  gallons or 5.73 drums.  This  translates to  an annual cost for disposal of approximately
 $10,309.09.

 Permit Fees~
        The Hudson/Sharp Central Impression Press used for this evaluation has two emission points, each
 requiring a permit from the New York State Departmentof Environmental Conservation (NYSDEC). The cost
 for these permits is dependent on the total tons per year emitted from each point. At 25 tons per year,  the cost
 is $100.  Over 25 tons, the cost goes up to $1,000.  For the current permits, Lustreprint was subject to the full
 $2,000 cost for the two emission points.  It is anticipated that, for the renewal of these permits, due to the use
 of water-based inks, that the cost will be reduced for an annual savings of $1,800 per year.

        It  should be noted that the corona treatment results in the generation and emission of ozone.  The
 equipment supplied by the manufacturer of the unit includes an ozone decomposer to limit emissions.   The
 permit issued  to Lustreprint for the ozone emission cojst $50 for an  estimated total of 34 pounds of ozone
 generated per year, assuming continuous operation of the unit. Once the second corona treater is installed, an
 additional permit will be required.                     '

        Thus total permitting costs would drop from $2000 at present to $300, once full implementation of water-
based inks  is achieved.
                                                 39

-------
        Water-based inks provide an advantage over solvent inks in that, when applied properly, they provide
more square feet of printing per pound of ink due to higher-solids content.  An additional advantage is the
reduced need of make-up solvent during the printing process.

        With a lower volatile content, water-based inks hold their color and viscosity for longer periods during
printing runs. Make-up additions, when necessary, are usually accomplished using the solvent identified as 8020,
which is 80% water and costs only $0.10 a pound versus an average of $0.57 for traditional solvent additives.

        The Lustreprint inventory provided information on the amounts of ink and solvent used for a nine month
period starting with March 1991. The report states that, on average, 1.27 pounds of solvent is'used for each
pound of solvent-based inks.  Relative to the amount  of make-up used per pound of solvent ink during the
evaluation, this amount appears excessive.  Lustreprint indicated that cleanup of the presses and equipment
subsequent to a fire at an adjacent facility during the study period may have impacted solvent use for this period.
Because of this,  all cost figures should be considered as estimates and any individual embarking upon this
technology is advised to do some preliminary tests within their own facility using the  type of equipment  they
project  to employ.

        A total of 15,030 pounds of water-based ink was used by Lustreprint for the nine months reported on
the ink  inventory. This represents 21% of the total 71,869 pounds of ink used in that period. At an average
savings  of $0.63 per pound, a savings of $9,468.90 was realized due to the substitution ($12,625.20 annualized).
If one extrapolates this annualized savings to a water-based ink usage factor of 100% (from 21%), the resulting
annualized savings would be $60,120.00.  The relative amount of various colors and types of inks can significantly
impact the projected saving, and may vary considerable for other printing operations.

        The cost differential between make-up for solvent- versus water-based ink is $0.47 per pound.  Using
the 1.27 pounds of solvent per pound of ink, the use of water-based ink without solvent make-up provided an
additional savings of $8,971 for the nine months ($11,96133 annualized).  Again, extrapolating to 100% water-
based ink would yield a savings of $56,958.71.

        The total annualized cost savings resulting from the substitution of water-based inks for solvent-based
inks would be $24,586.53 ($117,078.71 at 100%).

        It should be noted that even with conversion to 100% water-based ink, there would be some remaining
solvent. This solvent is present in the ink at a low level (around 20%).  Eventually, with the development of
solvent-free inks, this residual amount could be eliminated.   Once this is eliminated, other costs  could be
reduced, particularly those relating to hazardous material handling.

                                                  38

-------
        The most significant cost incurred by Lustreprint was the purchase of the corona treater. As discussed
                                                                                         I
previously, the treater, manufactured by Enercon Industries Corporation of Menomonee Falls, Wisconsin,
promotes the wetting and adhesion of water-based inks to the plastic web material.

        Other costs incurred included fans and duct work for enhanced drying capacity and air diaphragm pumps
for ink delivery to the press. New metering rolls were purchased for use with the high solids water-based inks.

Variable Costs                                                                           |

        The variable costs considered in this evaluation include ink and solvent costs, solid waste disposal costs
and permit fees. An increase of 4% (from 9.9% to 14%) in scrap material generated was noted over the course
of this study. However, this scrap rate was the same for both solvent- and water-based inks and was attributed
by Lustreprint to other factors such as training of new personnel As such, the cost for  scrap waste disposal is
not included as no appreciable increase in scrap material was attributed to the use of water inks.

Ink and Make-Up Solvent Costs-                    ;
        To determine the cost of ink used by Lustreprint, a computer printout of the ink inventory at Lustreprint
was obtained. The inventory includes information on each ink including a description, date received and last
used, amount on hand, current cost, and value of inventory stock. The current value provides an up to date cost
per pound for each ink. The costs are updated by Lustreprint on a regular basis. This allows Lustreprint to bill
on a replacement cost basis for inks used in printing.  (NJB.-this detailed computerized costing is considered
proprietary to Lustreprint and is not included in this report. The ink cost data described below was derived from
this report.)                                       ;
                                                   t
        The prices listed for the inks varied significantly for both water-based and solvent-based inks. Ink costs
are influenced by a number of factors such as ink type (Le., surface print inks versus laminating inks). Ink color
may also affect the cost as in the case of specialty colors, such as gold and silver, which are more expensive.

        To determine the cost of inks, an average cost j per pound of both solvent-based and waller-based inks
was calculated using the prices reported in the inventory provided by Lustreprint. It was determined that, on
average, solvent ink cost, $2.78 per pound.  The cost of water-based inks averaged out at $2.15 per pound
resulting in a cost savings of $0.63 for each pound of water-based inks used. Lustreprint noted that the bulk of
water-based ink purchases were for water white ink. White ink, for both water-based and solvent-based ink, is
the lowest cost color ink.
                                                  37

-------
                         TABLES. ECONOMIC FACTORS REVIEWED
                                          4 •.:-.;    -\. »

       FIXED/START-UP COSTS

       Hudson/Sharp C.I.
        Printing Press Upgrade
              Dryer/Ventilation
              Pumps
              Application Rollers
              Corona Discharge Treater
                     Purchase
                     Installation
              Basis Water Ink Inventory
              VARIABLE COSTS
                     Inks Cost (per gallon average)
                     Waste Disposal Calculated
                     Labor Variance
                     Off-Spec Product (Nonhazardous waste)
                     Utilities
              OVERHEAD
                     Insurance
                     Regulatory Compliance
                       DEC
                       OSHA
                     Training


Fixed Costs


       Lustreprint purchased equipment and completed a number of modifications to the Hudson/Sharp 48

Central Impression press in  order to accommodate the use of water-based  inks.  The  equipment and

modifications, along with  the costs, are listed below:
                             TABLE 6. FIXED COSTS (INSTALLED)

       Exhaust Fans, Supply Fans                                           $15,881
       DuctWork                                                          L200
       Metering Rollers                                                     6,900
       Pumps (air operated diaphragm)                                        2^10
       Corona Treater                                                      25,370
               spare parts                                                   L560
               proportional  speed control                                     1,410
               remote control station                                           920
               LCD digital watt meter                                          230
               ozone decomposer                                            3,590
               ozone decomposer differential pressure gauge                       475
               exhaust blower                                               2,030
               hinged electrode assembly, shroud option                           965
       TOTAL                                                           $62,901


                                              36

-------
        It should be noted that much of this waste decrease can be attributed to other factors and is not related
to any type of ink used  The WRITE Program evaluation and the use of the waste generation form increased
awareness of press operators and provided a deterrence to waste generation. This induced press operators to
reuse solvent for additional cleaning or in solvent inks.

ECONOMIC ANALYSIS
                                                   i
                                                   i                                      i
        An economic analysis of the changeover from;solvent- to water-based  inks is included as part of this
project (Appendix B).

        Fixed, variable and overhead costs are affected by this substitution and are considered in this evaluation.
Fixed costs include the purchase and installation of; new equipment,  primarily  consisting of the Enercon
Industries Corona Discharge Treater.   Costs for the  replacement of equipment, ancillary to  the  central
impression cylinder press, such as pumps, dryer upgrade, ink pans, etc., are also included in the fixed cost
analysis.

        Variable cost adjustments include the premium paid, or reduced cost, for water-based inks. The costs
for all inks used during each of the one-week periods was calculated with an average cost per gallon calculated
for all inks used. It was anticipated that, with the premium paid for water-based inks, the cost per gallon would
be higher for the weeks utilizing the greatest amount of water-based inks. Disposal costs were calculated using
the amount of solvent waste ink generated in gallons and the most recent disposal cost figures provided by
Lustreprint.  (However, the cost of waste due to scrap product was attributed to a lack of familiarization with
the operation of the corona discharge treater by Lustreprint plant personnel and was not included  in the
economic determination. The quantity of this non-hazardous waste should decrease over time.) Other variable
costs which were considered include variations in labor hours and utilities.

        Overhead costs also play a role in determining the cost savings.  Items such as the time previously
expended for regulatory compliance, insurance costs, j employee equipment and safety training  and  OSHA
compliance were expected to be reduced as a result of the removal of hazardous waste from the shop floor.
These potential cost savings were estimated from existing figures where available.

        A list of all the economic factors reviewed and considered in the analysis is shown in Table 3.
                                                  35

-------
                                    TABLE 4.  VOC REDUCTION


Weekl
Week 2
WeekS
Week 4
Total
Ink (Ibs)
2363
2,254
2,252
1,237
Factored*
VOC (Ibs)
1,772
1,634
1,633
897
%
Water Ink
52.9
2Z5
0.0
55.6
Reduced
VOC (Ibs)
827.5
1,251.7
L571.5
509.0
%
Reduction
533%
23.4%
0%
433%
        * This is calculated by taking 72.5% of the total ink quantity.                   !

        As can be seen from the table, for Weeks 1 and 2, the VOC generation decreased in proportion to the
percentage of water-based ink used. A 52.9% water-based ink utilization resulted in a 533% reduction in VOC
emissions.  Similarly, in Week 2, a water ink utilization rate of 225% resulted in a VOC emission reduction of
23.4%.

        For Week 4, the corresponding reduction in VOC emissions was less significant, where a 55.6% water
ink utilization rate resulted hi only a 433% reduction hi VOC emissions. Total ink use for Week 4 was 1,237
pounds of combined water and solvent ink.  This amount is approximately half of what was used during the other
three weeks of the study. The number of different inks used in Week 4, however, is comparable to that which
was used in Week  1.  With the same number of ink changes  at the printing stations and with each change
requiring a cleaning prior to the addition of new ink, the amount of cleaning make-up solvent relative to total
ink use is expected to increase. Since the same solvent materials are typically used by Lustreprint for both water
and solvent ink, the significance of the contribution to VOC emissions by clean-up solvent increases, thereby
reducing the effectiveness of VOC reduction by water inks.

Discussion  of Results of Analysis on Solid Waste

        Historically, 315 gallons of solid waste is generated by Lustreprint each month.  This translates to
approximately 1.5 55 gallon drums or 424 pounds per week.  Printing operations during Week 1 resulted in the
generation  of 55.5 pounds of solvent ink waste while Week 2 generated 20.0 pounds of water-based ink waste.
Therefore,  the net result was an 87 percent decrease from normal in solid waste generation in Week 1 of the
study (to 55.5 pounds), a 95 percent decrease hi Week 2 (to 20.0 pounds), and 100 percent elimination of solid
waste generation in Weeks 3 and 4.
                                                 34

-------
      0)
      €0
     ID
      S
0)
      O)

      O
     h-
                               o
                               10
                                  13
                                  0)
                                  CO
                                                 o
                                                 "(f

                                                    '
                                                     CO
                                         CO

                                         2
                                         O)
                                         o>
                                         DC
                                         0)
                             _  o
                                  «s
                                   o>
                                                  o
                                                  CM
                                               -  O
                                   0)
                                   e
                                   0)
                                  a.
                                                           co
                                         LU
                                         H;

                                         CC
              o
              o
              00
ooooooooo
oooooooo

-------
       As shown previously on Table 4, for Week 3 when only solvent ink was used, VOC emissions, calculated
from the material accounting, equaled 69.8% of the total ink use for the period.  This value fits well with the
equation derived by linear regression from the historical and solvent Only data, where VOC generation of 72.5%
of total ink use is expected.

       Next, the data points for the WRITE program total ink are shown in comparison to the Historical and
Solvent Only data and linear regression

       If one proceeds to plot the VOC emitted in pounds versus the percentage of water inks in the total ink
usage, the relationship in Figure 6 is obtained.  (Note that the data point for Week 4, as noted later in this
discussion, contains data that would have been affected by a higher proportion of ink color  changes and thus a
higher proportion of cleaning solvent and has been excluded.)  The relationship of decreasing VOC emissions
with increasing water use can clearly be seen.

       If we attempt a linear regression based on the three weeks in which ink  usage was similar, we obtain
a line with the relationship:

        y = -14.06x  + 1570.2,
        where x = % water-based ink used and y = pounds of VOC emissions.

        The linear regression shown in Figure 6 with an excellent correlation coefficient of 0.99997 is thus
obtained. From this  data it is apparent that there would be expected a reduction of 14 pounds of VOCs emitted
for each 1% increase in water-based ink used from a base level of about 2250 pounds usage (Lustreprinf s typical
weekly usage during this study).

        The primary motive for the substitution of solvent inks was to reduce the generation of VOC emissions
from the printing process.  The use of water-based inks, which are lower in solvent content, is expected to
favorably impact point source and fugitive emissions at the facility.  From the data collected during the four
weeks of the evaluation, the utilization of water-based inlcs varied from 0% in Week 3 to as much as 55.6% in
Week 4. This information, along with the historical data provided a basis for comparison.
                                                  32

-------
co
g
CO
CO
o
o
      m
              (0
              .a
        •o


 o    I

 •«••    _!.
 f*\    fl3
2    o

.^    £
        Xo
        »

        E
        tu
        O
        O
                o
                o
                to
                        o
                        o
                        o
                        CM
                       O
                       O
                       10
O
O
o
                                          I
o
o
10
                                                  o
                                                  o
                                                  •o
                                                  CO
                                                  o
                                                  o
                                                  o
                                                  CO
                                                    o
                                                    o
                                                    IO

                                                              0}

                                                             3
                                                                o»
                                                                9
                                                                re
                                                          X»-
                                                          ^^™

                                                      "c  -1
                                                      o  >
                                                      *-  o
                                                              o
                                                             CO
                                                  o
                                                  o
                                                  o
                                                  o
                                                  o
                                                  to
                                                             •   «0
                                                             Q  >
                                                             _  o
                                                              g  «•
                                                             -  UJ
                                                              o  t
                                                              «  cc.
to
                                                                        at
                                     31

-------