&EPA
United States
Environmental Protection
Agency
Office of Research and
Development
Washington, DC 20460
EPA/600/R-94/026
September 1993
Onsite Solvent
Recovery
-------
-------
ONSITE SOLVENT RECOVERY
by
Arun R. Gavaskar
Robert F. Olfenbuttel
Laura A. Hernon-Kenny
Jody A. Jones
Mona A. Salem
John R. Becker
Joseph E. Tabor
Battelle
Columbus, Ohio 43201
Contract No. 68-CO-0003
Work Assignment No. 2-36
Project Officer
Ivars Licis
Waste Minimization, Destruction, and Disposal Research Division
Risk Reduction Engineering Laboratory
Cincinnati, Ohio 45268
RISK REDUCTION ENGINEERING LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
EPA/600/R-94/026
September 1993
Printed on Reaydad Paper
-------
r
NOTICE
This material has been funded wholly or in part by the U.S. Environmental Protection
Agency (EPA) under Contract No. 68-CO-0003 to Battelle. It has been subjected to the Agency's
peer and administrative review and approved for publication as an EPA document. Approval does
not signify that the contents necessarily reflect the views and policies of the EPA or Battelle; nor
does mention of trade names or commercial products constitute endorsement or recommendation
for use. This document is intended as advisory guidance only to solvent-using industries in
developing approaches to waste reduction. Compliance with environmental and occupational
safety and health laws is the responsibility of each individual business and is not the focus of this
document.
-------
FOREWORD
Today's rapidly developing and changing technologies and jindustrial 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 environ-
mental 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 jone of the products of
that research and provides a vital communication link between the researcher and the user
community. i
Passage of the Pollution Prevention Act of 1990 marked a significant change in U.S.
policies concerning the generation of hazardous and nonhazardous wastes. This bill implements the
national objective of pollution prevention by establishing a source reduction program at the EPA and
by assisting States in providing information and technical assistance regarding source reduction. In
support of the emphasis on pollution prevention, the "Waste Reduction Innovative Technology
Evaluation (WRITE) Program" has been designed to identify, evaluate, arid/or demonstrate new
|
ideas and technologies that lead to waste reduction. The WRITE Program emphasizes source
reduction and onsite recycling. These methods reduce or eliminate transiDortation, handling,
treatment, and disposal of hazardous materials in the environment. The technology evaluation
project discussed in this report evaluates new technology applications by rpeasuring performance,
the release of solvents into the environment, energy use, and the associated economics.
E. Timothy Oppelt, Director
Risk Reduction Engineering Laboratory
HI
-------
ABSTRACT
This study evaluated the product quality, waste reduction/pollution prevention, and
economic aspects of three technologies for onsite solvent recovery. The technologies were
(1) atmospheric batch distillation, (2) vacuum heat-pump distillation, and (3) low-emission vapor
degreasing.
The atmospheric and vacuum distillation units were tested on spent methyl ethyl
ketone (MEK) and spent methylene chloride (MC), respectively. Samples of spent, recycled, and
virgin solvents at two industrial sites were subjected to physical and chemical tests to determine
solvent quality- The quality of the recycled solvent was found to be acceptable for use in the
specific applications examined during this study. Significant waste reduction was achieved by
reducing the volume of spent solvent to distillation residue needing disposal.
The low-emission vapor degreaser (LEVD) is a fully enclosed alternative to convention-
al, open-top vapor degreasing. It was found to reduce air emissions by more than 99%, compared
to a conventional vapor degreaser of the same production capacity.
Compared to disposal, the atmospheric and vacuum distillation units reduced operating
costs significantly. The estimated payback period for these units was found to be less than 2
years. The LEVD reduced operating costs by reducing solvent losses and labor costs. The
estimated payback for this unit was approximately 10 years.
The cost estimates were based on a full range of considerations including hardware,
engineering, installation, operation, maintenance, contingency funds, and energy use. They did
not, however, include potential changes in liabilities or impacts due to regulations planned, or in the
process of being implemented.
This report was submitted in partial fulfillment of Contract Number 68-CO-0003, Work
Assignment 2-36, under the sponsorship of the U.S. Environmental Protection Agency. This report
covers the period from May 1, 1991 to September 30, 1993, and work was completed as of
September 30, 1993.
IV
-------
CONTENTS
NOTICE
FOREWORD ! Hi
ABSTRACT j iv
ACKNOWLEDGEMENTS . viii
.I
SECTION 1 | .
Project Description j 1
Project Objectives ! 1
Description of the Technologies 2
Atmospheric Batch Distillation 2
Vacuum Heat-Pump Distillation 4
Low-Emission Vapor Degreasing 1. 5
Description of the Test Sites | g
Summary of Approach !. 10
Product Quality Evaluation . . .! 10
Waste Reduction/Pollution Prevention Evaluation 11
Economic Evaluation '. 11
SECTION 2
Product Quality Evaluation I. 12
Onsite Testing !. 12
Atmospheric Unit { 12
Vacuum Unit i. . . 12
LEVD j. . 13
Laboratory Analysis — Physical Characterization t 14
Atmospheric Unit [ 15
Vacuum Unit i 15
LEVD ; •;;;;;;; ie
Laboratory Analysis — Chemical Characterization I 17
Atmospheric Unit [ 17
Vacuum Unit ! 17
LEVD i '.'.'.'.'.'.'.'.'. 19
Product Quality Assessment [ 19
.
SECTION 3
Waste Reduction/Pollution Prevention Potential ; 21
Atmospheric Unit i 21
Vacuum Unit i 21
LEVD ',.'.'.'.'.'.'.'.'.'.'.'.'. 23
Waste Reduction/Pollution Prevention Assessment 27
-------
SECTION -4
Economic Evaluation [[[ 29
Atmospheric Batch Unit .............................................. 29
Vacuum Unit [[[ 31
LEVD [[[ 33
SECTION 5
Quality Assurance ....................... *• ................. ............ 40
Onsite Testing [[[ 40
Laboratory Analysis ................................. ................ 41
Limitations and Qualifications .......................................... 41
SECTION 6
Conclusions and Discussion ............................................... 44
Atmospheric Batch Unit .............................................. 44
Vacuum Heat-Pump Unit ............................................. 45
Low-Emission Vapor Degreaser ......................................... 46
SECTION 7
References ................................................ .......... 47
APPENDIX A
Additional Features of the Three Technologies .................................. 48
APPENDIX B
Analytical Testing of Spent, Recycled, and Virgin Solvents ......................... 50
APPENDIX C
Calibration of FID Instruments ............................................. 54
APPENDIX D
LEVD Chamber Concentrations at End of Cycle ................................. 57
APPENDIX E
Features of the Conventional Degreaser Used as a Comparison ...................... 62
APPENDIX F
Economics of Atmospheric Unit ................... • ........................ 64
APPENDIX G
Maintenance Costs of the Vacuum Unit ....................................... 69
APPENDIX H
Economics of Vacuum Unit ..... .......................................... 70
APPENDIX I
Energy Requirement of LEVD .............................................. 75
APPENDIX J
-------
TABLES
Table 1. LEVD Cleaning Cycle 8
Table 2. Onsite Testing of Atmospheric and Vacuum Units 13
Table 3. Onsite Testing of LEVD 14
Table 4. Physical Characterization of Solvent Samples 16
Table 5. Chemical Characterization of Solvent Samples 18
Table 6. Performance Characteristics of Methylene Chloride . . . 19
Table 7. Waste Reduction by Atmospheric and Vacuum Units. 22
Table 8. Emissions from LEVD 24
Table 9. Major Operating Costs for Atmospheric Unit 30
Table 10. Major Operating Costs for Vacuum Unit 32
Table 11. LEVD Cycle Time Breakdown for Test Runs 34
Table 12. Operating Costs for Vapor Degreasing 36
Table 13. Precision of Solvent Analysis . . . 42
FIGURES
Figure 1. Atmospheric Distillation Unit 3
Figure 2. Vacuum Heat-Pump Distillation Unit 4
Figure 3. Low-emission Vapor Degreaser 7
Figure 4. End of the Cleaning Cycle for Run 1 25
Figure 5. Variation of LEVD Cycle Time for Steel 35
Figure 6. Variation of LEVD Cycle Time for Various Metals 39
VII
-------
ACKNOWLEDGMENTS
The U.S. EPA and Battelle would like to acknowledge the cooperation of Bob Burmark
of the Washington State Department of Ecology. The authors also wish to thank the following
people for their important contributions and support during this study:
Dave Townsend, Durr Automation, Inc.
Jason Valia, Durr Industries, Inc.
Jorg Isenbugel, Durr GmbH.
Wil Chamberlin, Durr Industries, Inc.
Roland Meyer, Vaco-Solv Chicago, Inc.
Pete Scantlebury, Finish Thompson, Inc.
John Marcotte, Cooper Industries, Inc.
Ray Stowers, Navistar International Transportation Corporation
VIII
-------
SECTION 1
PROJECT DESCRIPTION
This study, performed under the U.S. Environmental Protection Agency's (EPA) Waste
Reduction and Innovative Technology Evaluation (WRITE) Program, was a cooperative effort
between EPA's Risk Reduction Engineering Laboratory (RREL) and the Washington Department of
Ecology. The objective of the WRITE Program is to evaluate, in a typical workplace environment,
examples of prototype or innovative commercial technologies that have potential for reducing
i
waste. The results of the evaluation then are made available to potential users.
i
For each technology to be evaluated, three issues were addressed. First, it must be
determined whether the technology is effective. Because waste reduction technologies usually
involve recycling or reusing materials, or using substitute materials or techniques, it is important to
verify that the quality of the product continues to be satisfactory for the intended purpose.
Second, the wastes and releases generated by the old and new systems must be calculated and
compared. Third, the economics of the new technology must be quantified and compared with the
economics of the existing technology. It should be clear, however, that traditional methods of
calculating economics may not be sufficiently accurate to account for a list of benefits that
represent a growing impact on the cost of doing business. These benefits include worker safety,
morale and the resulting productivity, reduced company liability, reduced regulatory compliance
(cost of permits, testing, record keeping, and reporting), increasing costs of disposal, as well as
company reputation as good citizen and steward of resources. The quantification of these benefits
is more difficult. This project did not include the evaluation of these benefits.
This study evaluated three technologies for recovering and reusing waste solvent on
site. They were atmospheric batch distillation, vacuum heat-pump distillation, and low-emission
vapor degreasing. These technologies were selected because of their waste reduction/pollution
prevention potential, potential for wide applicability, and opportunity for testing.
PROJECT OBJECTIVES
The goal of this study was to evaluate three technologies
recover and reuse waste solvents. This study had the following critical objectives
that could be used to
-------
• Evaluate the product quality resulting from use of each technology
• Evaluate the waste reduction/pollution prevention potential of each technology
• Evaluate the costs/benefits (economics) of using each technology.
Direct comparisons among the three units were not an objective of this study, but rather, the
suitability of each technology to its respective application and the consequential environmental
considerations were evaluated.
DESCRIPTION OF THE TECHNOLOGIES
The three technologies evaluated are described below. In each technology category,
a specific unit offered by a specific manufacturer was tested. However, other variations of these
units (with varying capabilities) may be available from several vendors.
Atmospheric Batch Distillation
This is the simplest technology available for recovering liquid spent solvents. Units
that can distill as few as 5 gallons (gal) or as much as 55 gal per batch are available. Some of
these units can be modified to operate under vacuum for higher-boiling solvents (300°F or higher).
In atmospheric distillation, the spent solvent is subjected to controlled heating. The solvent boils,
turns into vapor, and leaves the still. Contaminants with higher boiling points (e.g., oil, solids) are
left behind in the still, as distillation residue or still bottoms, and can be collected at the end of the
batch. Typically, this residue is sent out for disposal. Distillate vapors are condensed to obtain the
recycled solvent. Contaminant components that have lower boiling points than the solvent or that
form an azeotrope with the solvent cannot be separated (without fractionation features) and may
end up in the distillate.
The atmospheric batch unit used in this study was a Model LS-55D," manufactured
by Finish Thompson, Inc." Figure 1 describes this 55-gaI capacity atmospheric unit. A disposable
polymer bag lines the still, and the spent solvent is poured into it. After the unit's lid is closed and
sealed, the unit is switched on. An encapsulated heating element (direct heat) distills the solvent,
which condenses and collects in a clean drum. Temperature settings can be varied according to
' Mention of trade names or products does not constitute endorsement for use.
-------
Contaminated
Solvent
h:
Condenser
Stillbag
Heated
Walls
Electric
Heat Source
Reclaimed
Solvent
Figure 1. Atmospheric distillation unit.
(Source: Finish Thompson, Inc.)
the solvent. A sight glass indicates when no further distillate is available. jThe unit is then shut off
I
and allowed to cool. Alternatively, a timer or a temperature controller is available as an option.
The still bag containing the residue is removed and disposed of as hazardous waste.
The unit has several safety features, including explosion-proof design. Two thermo-
stat controls ensure that the recommended heater temperature is not exceeded. A water flow
switch/interlock is used to shut off the heater if water supply to the condenser is interrupted. For
the configuration tested, uncondensed vapors escapes along the fill hose ;at the distillate drum as
the drum slowly fills. Generally, solvent users have a flammable storage: area that can meet the
requirements of this operation. Roof vents also have been used in other, similar situations.
However, additional safety requirements may be levied by insurers, such as an explosion-proof
roof. Appendix A describes some variations in design and operation of atmospheric units offered
by the same vendor to suit different applications. j
-------
Vacuum Heat-Pump Distillation
The unit tested. Model 040* is manufactured by Mentec AG* in Switzerland and
supplied in the U.S. by Vaco-Solv Chicago, Inc.* As shown in Figure 2, the configuration is similar
to that of a conventional vacuum distillation system except that the pump, in addition to drawing a
vacuum, functions as a heat pump. No external heating or cooling is applied. The heat pump
generates a vacuum for distillation and compresses vapors for condensation. The Model 040 used
in the testing is suitable for solvents with boiling points up to 80°C.
Condensate Trap
Vapor Filter
Air
Condenser
Distillate
'I'
Residue
Figure 2. Vacuum heat-pump distillation unit.
(Source: Vaco-Solv Chicago, Inc.)
As shown in Figure 2, the spent solvent is continuously sucked into the evaporator by
a filling valve. The vacuum, drawn generates vapors, which are sucked into the heat pump,
compressed, and .sent to the condenser. Temperature stabilizes automatically according to the
specific solvent characteristics and the ambient. The condenser surrounds the evaporator allowing
heat exchange between the cool spent solvent and the warm condensing vapors. Because of this
heat exchange, this vacuum unit uses 50 to 75% less energy than a conventional distillation unit.
* Mention of trade names or products does not constitute endorsement for use.
-------
The energy savings estimate is based on the power ratings of the heat pump (vacuum unit) and the
electric heater (atmospheric unit).
i .
The heat pump is a single-stage rotary vacuum pump, modified to operate in a solvent
atmosphere. The pump oil is a type that is insoluble in solvent. Solvent vapor entering the pump is
kept free of solid and liquid impurities by a vapor filter and condensiate trap. An overflow
protection device guards against foaming in the evaporator by releasing thelvacuum.
A continuous distillate is produced and can be collected in a dean tank or drum. The
residue at the bottom of the still can be intermittently drained for spent isolvents containing less
than 5% solids. For spent solvents with a higher solids content, continuous draining by means of a
discharge pump may be necessary. -.-.]-
i
This unit can be retrofitted into an existing operation for continuous recycling.
Alternatively, it can be used in a semicontinuous mode, whereby the inlet hose is dipped into a
drum containing spent solvent. The unit then draws solvent continuously put of the spent solvent
drum until it is empty. The recycled solvent is collected through the outlet hose in a clean drum.
The unit is offered in nonexplosion-proof as well as explosion-proof design (for flammables).
Appendix A describes some variations in design and operation of vacuum units offered by the same
vendor to suit different applications. |
Vacuum distillation generally improves process efficiency, leaving much less solvent
behind in the residue. Although most solvents have fairly good heat stability, the stabilizers in the
solvent can degrade during heating. Because the solvent boils at a lower temperature under
vacuum, repeated solvent recycling is less harsh on the stabilizers. Baking and incrustation of the
residue due to heating also are avoided. The continuous mode of operation!of this unit is especially
suitable for degreasing applications, in which continuous recycling of the (solvent maintains good
solvent quality at all times. This results in a lower heating-energy requirements, compared to a
.
solvent that becomes progressively contaminated. The addition of a continuous distillation unit to
a degreaser, as described below, can increase the throughput of the degreaser by reducing the
cleaning cycle time and reducing the frequency of the degreaser shutdown time for solvent
replacement or internal distillation cycles, which is a feature of the LEVD. 1 These benefits have to
j
be weighed against the added costs of purchasing and operating the extra hardware.
Low-Emission Vapor Decreasing
Low-emission vapor degreasing (LEVD) is a technology currently used in Europe,
where vapor degreasers are regulated as a point source. Until recently; LEIVD units were not
-------
available in the United States, where a variety of conventional open-top vapor degreasers (with or
without lids) are used instead.
Conventional vapor degreasing is a common method for cleaning machined parts, and
the workload is inserted in a tank containing solvent vapors. The vapors condense on the parts,
and the contaminants (e.g., oil) on the parts drain out with the condensate. The condensate drops
into a sump, which is heated to produce more vapors. Before reaching the sump, the condensate
passes through a water separator. This process continues until the workload reaches the same
temperature as the solvent vapor and no more solvent condenses on it. The workload is then
withdrawn and cooled. The solvent in the sump is used until excess contaminants accumulate in
the spent solvent. • The spent solvent is then discarded and fresh solvent is added. The spent
solvent can be distilled by conventional means and reused. Alternatively, a continuous distillation
unit (discussed above) can be directly hooked up to the sump so that solvent is continuously drawn
from the sump, distilled, and returned to the sump.
However, previous studies (Battelle, 1992a) on conventional open-top vapor
degreasers have showed that a large part of the solvent (more than 90% in some cases) is lost
through air emissions. Although vapor degreasers are required to have primary cooling coils
(tapwater cooled) and a certain freeboard height (typically 0.75 times the width of the opening or
higher), air emissions are considerable. Air emissions are mainly workload-related, caused either by
dragout of solvent on the workload itself (and subsequent vaporization) or by disturbance in the air-
vapor interface during entry and exit of the workload. Other sources of air emissions are convec-
tion and diffusion during startup, operation, idling, shutdown, and, to a small extent, equipment
leaks. Air emissions are a concern for metal finishers because many of the solvents used in vapor
degreasing have been targeted by the EPA in the 33/50 Program. Both EPA and Occupational
Safety and Health Administration (OSHA) regulations are becoming more stringent.
Improvements available for conventional vapor degreasers include increased freeboard
height, refrigerated coils, and covers that eliminate drafts and reduce diffusion.
In contrast, low-emission vapor degreasers (LEVDs) are completely enclosed, airtight
units. The unit tested in this evaluation was the Model 83S (Size 1) which is manufactured in the
United States by Durr Automation, Inc."
Figure 3 shows the operation of the LEVD. Approximately 1 hour before the shift
begins, a timer on the unit switches on the heat to the sump. The solvent in the sump is allowed to
reach vapor temperature. At this point, the vapor is still confined to the enclosed jacket around the
working chamber. The parts to be cleaned (workload) are placed in a galvanized basket, which is
Mention of trade names or products does not constitute endorsement for use.
-------
lowered by a hoist from an opening in the top of the unit into the workjng chamber. Loads can
range from 330 to 1100 Ib (of steel parts) in this model. The lid is shut,! the unit is switched on,
and compressed air from an external source hermetically seals the lid. The lid remains shut until
the full cleaning cycle is completed.
Legend
*• Desorption Stage
>_ Adsorption Stage
* Liquid Solvent
•Water
Figure 3. Low-emission vapor degreaser.
(Source: Durr Automation, Inc.)
Table 1 shows the various stages in a typical cleaning cycle! The table also shows
typical time settings recommended by the manufacturer for each stage, as; well as the actual time
settings used during the testing for this evaluation. i
i
The first stage in the cycle is an optional liquid solvent spray. This stage, which can
be preset at 10 to 180 sec, is used only in certain cases when excessive contamination on the
workload requires mechanical cleaning action. The second stage is the Tvapor-fill" stage, during
which solvent vapor is introduced into the working chamber for the first time. The vapor enters
from near the bottom of the chamber and rises through the workload to a preset height, 1 ft below
the lid of the chamber. Condensate from the parts is continuously drawn into a water separator
and returned to the sump. This stage continues until the chamber tempprature reaches 100°C.
Stage time varies depending on type and mass of workload. The unit is then considered to be in its
-------
optimum "degreasing" stage. The unit remains in this degreasing stage for a brief period (20 to
180 sec preset) for maximum cleaning.
TABLE 1. LEVD CLEANING CYCLE
Stage
Solvent heatup (once a day)
Solvent spray (optional)
Vapor fill
Degreasing
Condensation
Air recirculation
Carbon heatup
Desorption
Adsorption
Vendor-Recommended
Time Settings
Variable3
10-180 sec
Variable13
20-180 sec
1 20 sec
1 20 sec
Variable0
60 sec
60-240 secd
Times Set
for this Testing
Variable8
not used
Variable15
60 sec
1 20 sec
1 20 sec
Variable0
60 sec
240 sec
" Requires approximately 1 hour on days following overnight shutdown when sump solvent
temperature drops to 70°C. After weekend shutdowns, when sump solvent temperature drops
to 20°C, it may take 1.5 hours for solvent to reach vapor temperature. Time on unit allows
automatic heatup prior to beginning of shift.
b Varies depending on workload metal and mass.
0 Carbon heatup took approximately 22.5 min during testing.
d At 60 sec, if monitor shows that chamber concentration is above 1 g/m3, then the adsorption
stage proceeds to the full 240-sec stage. This sequence repeats if necessary.
The unit then enters a 2-min (preset) "condensation" stage, in which the solvent
vapors in the chamber are condensed out by a refrigerated cooling coil at the bottom of the
chamber. The next stage is a 2-mih (preset) "air recirculation" stage. In this stage, the air-solvent
mixture in the chamber is recirculated continuously through a chiller to condense out more solvent,
which goes to the water separator and then the sump. The unit then enters a "carbon heat-up"
stage (stage time varies depending on initial carbon temperature). In this stage, the chamber air is
heated by a compressor fan and passed through a carbon unit consisting of a series of activated
carbon fiber mats. As the carbon mats heat up, the solvent captured during the previous cleaning
cycle is released to the circulating air. The desorbed solvent is then condensed out in the chiller.
This stage continues until the last (downstream) carbon mat reaches 70°C. At this point, the cycle
-------
goes into a 1-min (preset) "desorption" stage to complete the desorption.! The vendor is designing
future modifications to the unit that would incorporate twin carbon beds to eliminate carbon heat-
up and desorption stages. <
The carbon is now ready to adsorb the solvent from the current cleaning cycle. In the
"adsorption" stage, the chamber air is recirculated in the reverse direction — first through the
chiller and then through the carbon. Most of the residual solvent vapor iri the cold air is adsorbed
! i
on the carbon. The time for this stage can be preset to anywhere between 1 and 4 min. When
this preset time is reached, a PID (photoionization detector) probe inside the working chamber
verifies that the chamber air has less than 1 g/m3 of solvent and signals the air compressor to
release the seal on the lid to end the cycle. If the chamber air still has more than 1 g/m3 of
solvent, the cycle goes back to the desorption stage. The entire cleaning cycle is programmed and
requires no operator attention other than for loading and unloading the workload. The LEVD
system has no solvent exhaust during the entire cycle, except for a very small amount at the end,
when the lid is opened.
Several safety features are built into the unit. When vapor in the outer jacket
increases beyond a certain pressure, a thermostat turns off the heat to trie sump to prevent more
vapor from being generated. Inside the chamber, even though the lid is hermetically sealed, the
vapor level is not allowed to rise above the working level (1 ft below the lid). If the vapor level
I
rises above the working level, a thermostat signals a valve to stop any more vapor from entering
the chamber.
i
The unit can also be operated as a distillation unit to clean the liquid solvent in the
sump. In the distillation cycle, the unit is turned on without any workload in the chamber. After
most of the solvent is converted to vapor, the residue in the sump is drained out. The vapors in
the chamber are then condensed in the chiller to recover the solvent. The Jife of the solvent is thus
extended without a separate liquid distillation machine. Appendix A describes some variations in
j
design and operation of enclosed vapor degreasing units offered by the same vendor to suit
different applications. ' j
i
i
DESCRIPTION OF THE TEST SITES I
The two liquid distillation units were tested at industrial sites where the units have
been purchased and are being used. The LEVD unit was tested at its manufacturing site:
The Model LS-55D atmospheric unit was tested at Navistar International Transporta-
tion Corporation's Plastics Division in Columbus, Ohio. The atmospheric unit is used by Navistar to
reclaim various solvents used in its paints division. For this evaluation^ a test batch of spent
-------
methyl ethyl ketone (MEK) was processed through the unit. MEK is used to clean the spray
painting lines between colors. The spent solvent is collected in 55-gal drums and taken to the
flammables storage area to be recycled. The recycled solvent is reused for the same purpose, and
the residue is shipped off as hazardous waste.
The Model 040 vacuum unit was tested at Cooper Industries' Belden Division in
Richmond, Indiana. The Belden Division manufactures wires and cables. The recovery unit was
tested on spent methylene chloride (MC). The methylene chloride is used for cold (immersion)
cleaning of wires and cables when markings (ink) are to be removed. Cooper has modified the
Model 040 unit, by adding an air-cooled heat exchanger to allow it to operate at rates faster than
those recommended by the manufacturer. When solvent is circulated through the unit at a rate
higher than that recommended by the manufacturer, the distillate does not get enough chance to
cool down and exits the unit at a higher temperature. This can lead to higher vapor losses at the
point where the outlet hose enters the recycled solvent drum. Cooper uses a retrofitted air-cooled
condenser to cool the distillate and reduce vapor losses.
LEVD technology has been in use in Europe for some time. However, this technology
has only recently arrived on the U.S. market. Therefore, the 83S (Size 1) unit used in this study
was tested at the manufacturing location in Davisburg, Michigan. Machined steel parts were
processed several times to test the performance of the unit. Perchloroethylene (PCE) was the
solvent used during this testing.
SUMMARY OF APPROACH
A Quality Assurance Project Plan (QAPP), prepared at the beginning of this study
(Battelle 1992b), describes the detailed approach and scientific rationale used to evaluate the three
technologies. Sections 2,3, and 4 of this report describe the actual testing conducted. Section 5
(Quality Assurance) of this report describes any deviations from the test plan in the original QAPP.
For each technology, the evaluation covered issues of product quality, waste reduction/pollution
prevention, and economics.
Product Quality Evaluation
The objective of this part of the evaluation for the two liquid distillation units was to
show that the recycled solvent was of sufficient quality to allow its reuse. The approach was to
process one 55-gal drum of spent solvent each through the batch and continuous units. Samples
of the spent and recycled solvents were taken and analyzed to determine the improvement in
10
-------
quality. A sample of virgin solvent also was collected at each site and subjected to the same tests
as a basis for comparison. . j
...... • j .
The objective for the LEVD was to process test batches of j machined parts, covered
with cutting oil, through the unit and to ensure that the unit was performing its basic function —
effective parts cleaning. • j
Waste Reduction/Pollution Prevention Evaluation
The waste reduction potential of the two distillation units was measured in terms of
the amount of solvent that could be recovered and reused (current practice), rather than sent out
i
for disposal (former practice). 1
The pollution prevention potential of the LEVD was measured in terms of its ability to
1
prevent atmospheric emissions of solvent. Air emission information for conventional degreasers
was obtained from previous EPA-sponsored studies (Battelle, 1992a, EPA, 1989). The overall
potential for applying the technologies was estimated on the basis of annual solvent usage in the
United States.
Economic Evaluation
For the two liquid distillation units, the economic evaluation consisted of a comparison
between the cost of disposal versus the savings through recycling. For the LEVD, the cost of the
l
LEVD operation was compared to the cost of operating a conventional degreaser of similar
production capacity.
11
-------
SECTION 2
PRODUCT QUALITY EVALUATION
This section describes the testing that was done to ensure that the new technology
provides an acceptable product — clean solvent or clean parts.
ONSITE TESTING
The tests conducted at the evaluation sites are described below. All solvent samples
collected were refrigerated until analysis.
Atmospheric Unit
Table 2 describes the samples collected and data gathered during the test run.
Samples were collected in precleaned disposable glass tubes. The still was lined with a still bag,
and the spent methyl ethyl ketone (MEK) from the test drum was pumped into it. The lid was
closed and sealed, the tapwater flow was started through the condenser, and the unit was
switched on. It took about 1.5 hr for the first condensate to appear through the outlet hose into
the clean drum; this time is typical for the operation. The drum used to collect the recycled solvent
was an empty drum that once had contained virgin MEK. The site reuses its empty drums in this
fashion. The volumes obtained for the run were verified against the operator's logs to compare the
test numbers with those from previous runs. The test run was determined to be typical of the
operation.
Vacuum Unit
Table 2 describes the samples and data collected during the testing of the vacuum
heat-pump distillation unit. It was noted that the sample collected as a "virgin" sample was
actually a sample of methylene chloride (MC) that was obtained by Cooper from a solvent recycling
company. The characteristics of this "virgin" solvent meet the requirements for the company's
application, and this solvent has been used satisfactorily at the site in the past.
12
-------
TABLE 2. ONSITE TESTING OF ATMOSPHERIC AND VACUUM UNITS
Recycling
Unit
Atmospheric
Vacuum
Solvent
Type
MEKa
MCb
Spent
Solvent
(gal)
55
55
Recycled
Solvent
(gal)
39
48
Residue
(gal)
16°
3d
Run Time
(hrs)
12
12
Samples
Collected
j Spent
Recycled6
| Virgin
Spent
! Recycled8
! Virgin
Sample
Number
MEK-S
MEK-R
MEK-V
MC-S
MC-R
MC-V
a MEK = methyl ethyl ketone
b MC = methylene chloride
c No significant losses to the atmosphere.
d
Approximately 4 gal of solvent per 55-gal batch are lost due to air emissions.
Duplicate analysis performed on recycled samples [(MEK-R Dup) and (MC-R Dup)].
I
To speed up the processing rate, the vacuum unit was being operated at a faster rate
than that recommended by the manufacturer. The unit's built-in condenser-evaporator heat
exchange was not sufficient for this rate. Cooper had therefore attached an additional air-cooled
condenser at the outlet. With this extra device, the vapor loss at the outlet was restricted to
approximately 4 gal per 55 gal of spent solvent. To prevent the release of this vapor into the work
area, the vapor was led through a pipe to the roof of the facility and discharged at a rate within
I
applicable State regulations. According to the manufacturer, this vapor loss can be prevented by
operating the vacuum pump at a slightly lower rate. i
LEVD
Table 3 describes the testing conducted on the low-emission vap>or degreaser (LEVD)
using perchloroethylene (PCE) solvent. Runs were conducted on machined steel parts. The parts
were steel ridged cylinders (approximately 2-in diameter x 5-in height) and 6-in-diameter recessed
steel disks. Runs were started variously with and without cutting oil on the parts. Total cycle
times were recorded for all completed runs. Runs 1, 4, and 7 were interrupted for the reasons
mentioned in Table 3. Because the same batch of parts was used for each run, parts were either
cold (ambient) or hot (see Table 3) depending on the cooling time allowed between runs. The total
cycle time varies depending on the initial temperature of the parts. If parts are already hot, the
total cycle time is reduced. In general, the runs that start out with cold parts are more reflective of
actual operating conditions. The addition of oil to the parts did not greatly affect the total cycle
13
-------
time. The total cycle time varied considerably depending on the workload mass. The masses of
the steel parts given in Table 3 do not include the mass of the basket, which weighed 65 Ib.
TABLE 3. ONSITE TESTING OF LEVD
Day
No.
1
1
1
1
2
2
2
2
2
Run
No.
1b
2
3
40
5
6
7d
8
9
Initial Parts
Temperature
Ambient
Hot0
Ambient
Hotc
Ambient
Hot0
Hot0
Ambient
Hot0
Weight of
Steel Parts" (Ibs)
165
165
900
160
165
165
165
915
160
Oil on Parts
at Start
No
No
No
No
Yes
Yes
Yes
Yes
No
Total Cycle
Time (min)
_b
39
67.5
_d
50.5
40
_8
69
40
0 Not including the weight of the 65-lb basket.
b Run 1 was interrupted because of power outage due to stormy weather.
Was continued after machine was reprogrammed.
0 Approximately 60 to 70 °C above ambient.
d Purpose of Run 4 was to collect liquid solvent during the condensation and desorption cycles.
This run was not timed.
0 Run 7 had to be abandoned at carbon heatup stage due to carbon overheating. Wiring
problem was corrected and normal functioning was restored for the next run.
In all the runs starting with parts dipped in cutting oil, the cleaned parts were visually
examined for traces of oil contamination. This was done to ensure that the unit was performing its
basic function, degreasing. No oil contamination was noticed on the surface of the parts from any
of these runs.
LABORATORY ANALYSIS - PHYSICAL CHARACTERIZATION
Samples collected during onsite testing of the atmospheric and vacuum units were
refrigerated and sent to the laboratory for analysis. The laboratory tests performed on the solvent
samples included:
14
-------
• color (ASTM D1209, D2108)
• specific gravity (ASTM D891-89, D2111-85)
• absorbance (EPA Method 110.3)
• nonvolatile matter {ASTM D1353, D2109)
• conductivity (EPA 120.1)
• water content (ASTM D1364, D3401)
« acidity (ASTM DT613)
• purity (ASTM D2804).
For the MC samples, the following additional tests were conducted:
• pH (ASTM D2110) .
• acid acceptance (ASTM D2942)
• corrosion (ASTM D2251).
Appendix B describes the laboratory tests conducted on the solvent sampl
the standard methods used for the analysis. Duplicate analyses were
samples from both sites.
Atmospheric Unit
Table 4 describes the results of the physical characterization
samples. The spent solvent appeared to have collected a lot of dissolved
contamination during its use as a cleaner for paint lines. The recycled
similar in appearance and color to that of the virgin sample. The specific
recycled samples fell between the values of the spent and virgin samp es.
recycling had caused a definite improvement, although not quite to vii
measurements indicated sharp differences between the spent solvent
There was very little difference between absorbance measurements on
samples. Therefore, appearance, color, specific gravity, and absorbam
indicators of solvent quality for onsite operators.
s, their significance, and
p< rformed on the recycled
Vacuum Unit
Table 4 describes the results of the physical characterizat on
samples. The spent MC differed noticeably from the recycled and "virgin
and color. The spent solvent appeared to have collected a substantial
primarily in solution, during its use for removing marking ink from wires. Specifi
15
of the atmospheric unit
and sedimentitious
samples were relatively
gravity value of the
This indicated that
gin grade. Absorbance
the recycled solvent.
the recycled and virgin
e could serve as quick
and
of the vacuum unit
" samples in appearance
amount of contamination,
ific gravity and
-------
TABLE 4. PHYSICAL CHARACTERIZATION OF SOLVENT SAMPLES
Sample
Atmospheric
Unit (MEK)
Spent
Recycled
Recycled Dupb
Virgin
Vacuum Unit (MC)
Spent
Recycled
Recycled Dupb
Virgin
Appearance
Dark gray
w/sediment
Clear
Clear
Clear
Dirty gray-
brown
Clear
Clear
Clear, tinge
of yellow
Specific Absorbance
Color8 Gravity 400 nm 500 nm 600 nm
- c 0.845 4.5 4.5 4.5
5 0.827 0.014 0.000 0.000
5 0.821 0.016 0.002 0.001
5 0.800 0.004 0.007 0.005
-° 1.220 1.824 1.674 1.435
5 1.286 0.010 0.001 0.004
5 1.288 0.014 0.011 0.006
10 1.298 0.010 0.010 0.004
* On a scale of 5 to 500, with 500 being the darkest color.
b Duplicate analysis of the same sample.
0 Not comparable with standards. Sample was too dirty.
absorbance characteristics of the recycled samples had been restored to levels fairly close to those
of the "virgin" sample.
LEVD
Other than onsite measurements made during testing of the LEVD, there were no
samples or laboratory analyses.
16
-------
LABORATORY ANALYSIS - CHEMICAL CHARACTERIZATION
Chemical characterization parameters and the standard methods
Appendix B. Duplicate analyses were performed on the recycled samples
used are described in
from both sites.
Atmospheric Unit
Table 5 describes the results of the chemical characterization. Nonvolatile matter
(contamination) accounted for nearly 7% of the spent MEK sample, j This was reduced to
approximately 0.002% in the recycled sample, a value indistinguishable from that of virgin grade.
Conductivity and acidity values of the recycled samples fell between those of the spent and virgin
samples, indicating some improvement in these parameters. The water! content increased from
approximately 1.9% in the spent sample to approximately 5.5% in the recycled samples. This
indicates that water contamination present in the spent solvent transfers to the distillate.
However, the fact that the volume of the distillate is roughly 30% lower jthan the total volume .of
the initial spent batch explains only about 15% of this increase in water concentration. The
remaining water must have entered the recycled solvent during the recycling process itself,
possibly due to a slight leakage from the water-cooled condenser. i
I
The MEK purity of the recycled sample showed a substantial improvement from the
spent sample, increasing from 78% to about 85%. The large decrease in nonvolatile matter during
j
recycling (discussed above) accounts for most of this increase in purity. Of the 15% impurity in
the recycled sample, 5% is water as discussed above. The remaining 10% impurity is probably
due to the codistilling out of paint thinner solvents (proprietary blends) presenT; in the spent solvent
and having similar boiling points. j
Vacuum Unit
Table 5 describes the chemical characterization of MC samples from the vacuum unit
testing. Caution should be exercised in interpreting the results because the so-called "virgin"
solvent was actually purchased by the site from a solvent recycling company. Over time, this
"virgin" solvent grade has proved sufficient for its application — removing Imarking inks from wires
and cables. !
17
-------
TABLE 5. CHEMICAL CHARACTERIZATION OF SOLVENT SAMPLES
Sample
Nonvolatile
Matter
(mg/1 00 mL)
Conductivity
(//mhos/cm)
Water
Content
(% by wt.)
Acidity
(as acetic
acid wt. %)
Purity
{% by wt.)
Atmospheric Unit (MEK)
Spent 6,951
Recycled 2.6
2.0
Recycled
Dup
Virgin
2.2
7.05
3.30
3.40
1.15
1.89
5.42
5.56
0.09
.055
.003
.003
.002
78.41
85.02
85.54
99.09
Vacuum Unit (MC)
Spent
Recycled
Recycled
Dup
Virgin
34,101 ppm
20.37 ppm
17.88 ppm
57.16 ppm
1,063
137
136
36
0.27
0.25
0.24
0.14
NA
NA
NA
NA
NA
86.4
NA
90.1
NA « Not Applicable
Nonvolatile matter (contamination) content dropped from 3.4% in the spent to
approximately 0.002% in the recycled samples. Conductivity values of the recycled samples were
indistinguishable from the value of the "virgin" solvent but were clearly lower than that of the
spent solvent, indicating an improvement. Water content values of the recycled samples were
between those of the spent and the "virgin," without any of the values being particularly high.
The MC purity of the recycled solvent was 86%. This compares favorably with the
"virgin" sample purity of 90%, given that much of the 10% impurity in the "virgin" sample is
probably volatile. Volatile impurities present in the "virgin" solvent will codistiil out into the
recycled solvent.
18
-------
Table 6 describes some performance characteristics of MC
{a halogenated solvent).
The pH of the water extract of the recycled solvent was fairly close to the "virgin" value of 7.
Substantial improvement is indicated from the spent sample's pH of 5. The lower pH indicates the
presence of potentially corrosive components in the spent solvent. The acid acceptance for all
samples was below 0.16% NaOH, the minimum level specified for vapor degreasing-grade MC.
Because the "virgin" sample also was below this level, it is difficult to judge the effect of recycling
on stabilizer content. The corrosion test on steel and aluminum yielded noticeable corrosion only in
the case of the steel strip placed in the spent solvent sample. No such corrosion was evident due
to the recycled solvent, indicating that recycling improved the quality.
TABLE 6.
Sample
Spent
Recycled
Recycled Dup
Virgin
PERFORMANCE CHARACTERISTICS OF METHYLEf
pH
5.0
6.7
6.6
7.0
" Measured as equivalent NaOH wt%.
b In terms of oxidation products (rust)
c Steel strip showed a number of rust
Acid
Acceptance" Steel
0.032 Rustc
0.004 No rust
0.005 No rust
0.003 No rust
showing up on the metal surface.
marks on both sides.
IE CHLORIDE
Corrosion
Aluminum
No rust
No rust
No rust
No rust
LEVD
Only onsite measurements were made during the testing of [the LEVD. No samples
were taken or laboratory analyses performed. j
PRODUCT QUALITY ASSESSMENT
Substantial improvement in the quality of the spent MEK solvent feed was brought
about by the atmospheric unit. Standard solvent quality-monitoring parameters such as specific
19
-------
gravity, absorbance, and conductivity showed a relatively good agreement between recycled and
virgin solvent values. Nonvolatile matter was virtually removed from the recycled solvent. The
MEK purity rose from 78% in the spent solvent to 85% in the recycled solvent. The recycled value
was lower than the >99%-pure virgin grade, but the impurities are expected to be mainly paint
thinners (solvents), which do not interfere during reuse in cleaning spray painting lines between
colors. The presence of 5% water (traced to a small leak in the water-cooled condenser) in the
recycled solvent is of concern, although the site did not report any adverse effect on the painting
operation. The distillation unit has been at the site for several years and its condenser coil needs to
be changed.
The vacuum unit considerably improved the quality of the spent MC solvent feed. The
evaluation was complicated by the fact that the "virgin" solvent was actually purchased by the site
from a solvent recycling company. This "virgin" grade itself was analyzed to be 90% pure MC.
Hence, a large component of the spent solvent feed to the vacuum unit was (probably volatile)
impurities accumulated over several previous recycles at the site from other sources. Even then,
an 86% recycled MC purity was achieved by the vacuum unit. Nonvolatile matter was virtually
removed. Standard solvent quality monitoring parameters, such as specific gravity, absorbance,
and conductivity, showed relatively good agreement between recycled and "virgin" values,
indicating that the recycled material was a fairly close approximation of the "virgin" grade. The pH
of water extracts from the spent, recycled, and virgin samples showed that both the recycled
solvent and the "virgin" sample had a pH close to 7, compared to a potentially corrosive spent
sample pH value of 5. The corrosion test showed that spent solvent caused corrosion of steel,
whereas the recycled solvent caused no noticeable corrosion.
The performance of the LEVD enabled oily parts to be cleaned to acceptable levels. It
was noted that the enclosed design has potential for reducing water contamination of solvent from
at least one source. Water contamination during degreasing normally occurs either through parts
(workload) coated with water-soluble machine oils or through condensation of atmospheric
moisture. In open-top vapor degreasers, there is a tendency for atmospheric moisture to condense
out on cooling coils (especially refrigerated coils). In the enclosed design, only a limited amount of
ambient air (equal to the free volume of the loaded chamber) comes into contact with the cooling
surfaces of the unit. Therefore, there is less atmospheric contribution of water to the liquid solvent
in the LEVD. Lower water contamination implies lower depletion of solvent stabilizers and acid
formation by hydrolysis.
20
-------
SECTION 3
WASTE REDUCTION/POLLUTION PREVENTION POTENTIAL
The waste reduction/pollution prevention aspects of the: three technologies are
discussed in this section.
i
|
ATMOSPHERIC UNIT j
Table 7 describes the waste reduction achieved at the test site by this technology.
Annual solvent usage was obtained from records maintained at the site; during the year prior to
recycling. Annual waste generation numbers were estimated from the data in Table 2 and from
site records. The site generates 880 gal/yr of spent methyl ethyl ketone [MEK). With the disposal
option, this entire volume (along with its 55-gal containers) would havb been disposed of as a
Resource Conservation and Recovery Act (RCRA) hazardous waste through a waste disposal
company. Through recycling, this waste has been reduced to 262 gal of distillation residue {plus
its containers), which is disposed of as RCRA hazardous waste. I
The recycled solvent leaving the outlet hose has a near-ambient temperature, and
vapor losses during recycling are minimum. Previously, the manufacturer supplied an airtight outlet
hose-to-drum connection and a pipe to vent vapors to the roof of the facility. However, this has
been discontinued, and the outlet hose now hangs loosely into the drum. The recycling unit is
operated in an area of the site reserved for flammables storage. j
F ' "
MEK is a hazardous chemical listed on the Toxic Releases Inventory (TRI). It is also on
the EPA's list of 17 chemicals targeted for 33% reduction by 1992 and 50% reduction by 1995.
i
I
VACUUM UNIT !
Table 7 describes the waste reduction achieved at the test; site by this technology.
The annual solvent usage was obtained from records maintained at the site during the year prior to
recycling. The annual waste generation numbers were estimated from the i data in Table 2 and from
site records. The site generates 3,000 gal/yr of spent methylene chloride (MC). With the disposal
21
-------
TABLE 7. WASTE REDUCTION BY ATMOSPHERIC AND VACUUM UNITS
-Disposal Option - - Recycling Option -
Wastestream Annual Volume Wastestream Annual Volume
Atmospheric Unit Test Site:
Spent MEK 880 gal
Drums 1 7 drums
Vacuum Unit Test Site:
Spent MC 3,000 gal
Drums 55 drums
Distillation Residue
Still bags
Cooling water
Drums
Distillation residue
Air emission
Drums
Used oil
262 gal
17 bags
1 8,360 gal
5 drums
136 gal
218 gal
3 drums
1 gal
option, this entire volume (plus its 55-gal containers) would have been disposed of as a RCRA
hazardous waste through a waste disposal company. Through recycling, this waste has been
reduced to 136 gal of distillation residue (along with its containers), which is disposed of as RCRA
hazardous waste. A very small sidestream of used oil is generated through a routine oil change on
the vacuum pump. This oil is combined with other waste oil generated on the site and is disposed
of according to State regulations for used oil disposal.
According to the manufacturer, air emissions due to the recycling process itself are
largely avoidable, provided that the operating procedures recommended by the manufacturer are
followed. However, this site has modified the unit to process faster, and this results in some air
emissions (218 gal/yr based on Table 2) due to incomplete condensation of the vapors. These air
emissions are vented from around the outlet hose to the roof through a temporary pipe in
accordance with applicable regulations.
22
-------
MC is a hazardous chemical listed on the Toxic Releases Inventory (TRI). It also is on
the EPA's list of 17 chemicals targeted for 33% reduction by 1992 and 50% reduction by 1995.
LEVD
The pollution prevention aspect of the low-emission vapor degreaser (LEVD) was the
main focus of this technology. The enclosed design of the working chamber restricts emissions to
two paths. Normally emissions occur when the entire cleaning cycle is ,complete. At this point,
the lid is opened by discontinuation of the compressed air pressure that seals the lid, and retracted
to uncover the chamber. The opening releases solvent vapor that has not been evacuated from the
i
chamber during the condensation or adsorption stages of the cycle. The other potential source of
air emissions is leakage of equipment seals, valves, or piping.
Table 8 shows the emissions recorded from the LEVD. The approach used was to
insert a flame ionization detector (FID) probe into the working chamber below the designated vapor
level. Measurements with this FID probe were started during the adsorption stage and continued
through the end of this stage, until after the lid was released and opened. The objective was to
measure the concentration of any residual perchloroethylene (PCE) just before releasing the lid and
after the lid is retracted. A second FID probe was used to take continuous measurements all
around the unit during operation, with special emphasis around the lid 1:0 check for leaks. Both
FIDs were calibrated with PCE standards (see Appendix C). The outputs from both FIDs were
separately channeled to a data capture system. The measurements were [recorded on both a chart
recorder and a computer readout. The ambient levels in the indoor facility on the test days were
consistent at around 3 to 4 ppm; this low level may be due to exhausts from forklift vehicles that
I
were operating in the facility. i
Figure 4 shows how a typical cleaning cycle ends on this unit. (See Appendix D for
figures showing other runs.) Figure 4 shows the measurements taken (luring Run 1. The same
pattern was evident in the other runs. In Figure 4, time zero corresponds; to the start of measure-
ments in the working chamber. The FID probe in the working chamber was not activated before
this point to prevent large concentrations of solvent vapor from being drawn into the FID. The FID
had been calibrated for accuracy in the 0 to 1,000 ppm range; particular care was taken in the
range below 1 g/m3 (approximately 150 ppm of perchloroethylene), the concentration above which
the lid normally is not allowed to be opened by a programmed controller. At time zero, the
chamber FID probe was in place in the working chamber, and the ambient FID probe was
positioned outside the unit, near the lid seal. j
23
-------
TABLE 8. EMISSIONS FROM LEVD
Run No.
1
2
3
4
5
6
7
8
g
Target0
Final Chamber
Concentration0
(ppm)
52
75
92
NMd
43
47
NMd
78
NMd
150
Total PCE
Emission15
(Ib/cycle)
0.0005
0.0007
0.0008
—
0.0004
0.0004
—
0.0007
—
0.0013
Total
Cycle Time
(min)
—
39
67.5
—
50.5
40
—
69
—
60°
Emission Rate
(Ib/hr)
—
0.0011
0.0007
—
0.0005
0.0006
—
0.0006
—
0.0013
a At the moment when the seal on the lid is released.
b Based on 150 ppm = 1g/m3 of PCE and a chamber volume of 0.6 m3.
0 Normally the machine is programmed to release the lid when solvent concentration in the
chamber falls below 1g/m3 (150 ppm of PCE). This target was easily met in all the test runs.
d Not measured due to difficulties in completing the run as described in Table 3.
8 Expected cycle time for 560 Ib of steel parts (workload).
As seen in Figure 4, the chamber concentration began dropping sharply as the 4-min
(240-sec) adsorption cycle started. Throughout the adsorption cycle, the ambient probe just
outside the lid indicated concentrations between 3 and 4 ppm, which were equal to the indoor
ambient concentrations in the test facility at locations far from the LEVD unit. Just before the
adsorption cycle ended, the chamber FID was reading 52 ppm, well below the targeted 150 ppm.
After 240 sec, the compressed air pressure sealing the lid was automatically released, but the lid
remained completely in place. At this point, the chamber concentration began showing some
fluctuations as the air in the chamber was disturbed, but it stayed within a 40- to 50-ppm range.
A few seconds later, the lid was retracted completely to allow the chamber air full access to the
24
-------
c
tt
u
u
D)
as
JU
o
0)
-C
•u
UJ
3
O)
3N31AH13OaO1HOH3d SV Wdd NOI1VH1N3ONOO
25
-------
ambient. At this point, the chamber concentration dropped sharply as the residual solvent vapor in
the chamber dispersed. The ambient FID probe, which was near the chamber opening, showed a
corresponding increase (to 6 ppm). Both FIDs soon stabilized to facility ambient levels (3 to 4
ppm).
Later, as the basket of cleaned parts was raised out of the chamber, the second FID
probe was thrust into the basket and near the parts. No elevated readings above ambient were
sensed near the parts, indicating that the parts were free of solvent.
During the entire operating cycle, for five of the runs (Runs 1, 5, 6, 8, and 9),
a continuous perimeter survey to check for leaks was carried out around the unit. Special attention
was given during the vapor-fill stage to the edges of the lid, where the top opening of the working
chamber is sealed. At times during each run, the front and rear panels, which loosely surround the
unit components, were removed to allow probing around the valves, piping, seals, etc.
At two points during the perimeter survey, solvent vapor was detected above facility
ambient levels. The causes were identified as external to normal unit operation, and were
remedied. During Run 5, a slight elevation in PCE concentration was noticed when a front panel
was removed, and the second FID probe was held at the point where a temporary opening had
been drilled in the chamber wall to insert the chamber FID probe for this testing. The seal around
the chamber probe was tightened, and the elevated reading returned to ambient. During the vapor-
fill cycle of Run 8, as the FID probe was passed around the rear of the unit near the lid, an elevated
reading was noticed. It momentarily jumped to more than 100 ppm, stayed at around 30 ppm for
30 sec, and then tapered off. When the pressure gauge on the facility's compressed air supply
was checked, it showed that the air compressor had malfunctioned, causing the compression seal
around the lid to be released temporarily. When the compressed air supply returned to normal, the
elevation in the reading dropped back to ambient.
The above testing indicates that the only significant emission from the LEVD is at the
end of the cycle when the lid is opened and retracted. Table 8 shows the final chamber concentra-
tion (just before the lid is opened) for several of the test runs. In all the test runs, the solvent
concentration was well below the 1 g/m3 (150 ppm PCE) targeted. Therefore, 1 g/m3 can be taken
as the achievable concentration. The volume of the working chamber is 0.6 m3. Assuming that all
the residual solvent vapor (1 g/m3) in the chamber has been discharged to the ambient area, the air
emission through the top opening is 0.21 g (0.00132 Ib). Therefore, the typical discharge of
solvent through air emission is 0.00132 Ib/cycle. From Section 4, if 560 Ib of oiled steel parts are
processed, the cleaning cycle would take 1 hour. Therefore, the air emission from this LEVD mode
that processes 560 Ib of steel parts/hr is 0.00132 Ib of solvent/hr.
26
-------
As described in Appendix E, a typical conventional open-top vapor degreaser that can
clean at a similar rate (approximately 560 Ib of steel parts/hr) would have a 4.5 ft2 opening. This
conventional degreaser typically would emit 0.147 Ib of solvent/ft2/hr {Appendix E), or 0.662 Ib of
solvent/hr, from the 4.5-ft2 opening during continuous operation. !
Therefore, the LEVD reduces air emissions by more thani 99% compared with air
emissions from the typical conventional open-top vapor degreaser (i.e.! having a 0.75 freeboard
ratio, primary cooling coil, electric hoist, and no lip exhausts) used in this calculation.
The OSHA exposure limit for PCE is 25 ppm for an 8-hr TWA (time-weighted average).
Personal air sampling (in accordance with OSHA guidelines) was not conducted during this
evaluation, but the 3- to 4-ppm PCE levels near the unit at all times during operation and the 6-ppm
(or lower) levels at the edge of the chamber opening for about 2.5 mirj when the lid is retracted
completely (Figure 4) are well under the OSHA exposure limit. !
* !
WASTE REDUCTION/POLLUTION PREVENTION ASSESSMENT I
The atmospheric unit evaluated in this study has good solvent waste reduction
potential compared to disposal. At the test site, annual hazardous solvent waste generated for
I
disposal was reduced from 880 gal to 262 gal. A sidestream consisting Of 18,360 gal/yr of cooling
water (tapwater) is generated during recycling, but should not be a problem because this tapwater
does not come into contact with any solvent contamination and can be reused elsewhere as
process water. The user should be careful that there are no leaks iri the cooling water lines,
through which the water could be contaminated with solvent or vice versa.
i
The vacuum unit evaluated in this study has good solvent iwaste reduction potential
compared to disposal. At the test site, annual hazardous solvent waste generated for disposal was
reduced from 3,000 gal to 136 gal. At this test site, a largely avoidable' sidestream of 218 gal of
air emissions also was generated during recycling; the sidestream can be avoided by operating the
vacuum pump at a slower rate. j
The LEVD reduced air emissions by more than 99% compared with the estimated air
I
emissions from a similarly sized conventional open-top vapor degreaser (with a 0.75 freeboard
ratio, primary cooling coil, electric hoist, and no lip exhausts). Even allowing for equipment- and
operation-related factors and auxiliary controls (e.g., increased freeboard ratio, refrigerated coils,
covered opening, reduced room draft/lip exhaust velocities) used to reduce air emissions from a
conventional degreaser, the pollution prevention performance of the completely enclosed design
used in this study would be difficult to match. The perimeter ambient survey with a FID probe
indicated that the OSHA exposure limit of 25 ppm (8-hr TWA) would be met by the LEVD.
27
-------
The pollution prevention potential of this unit is further enhanced by its ability to
perform as a liquid solvent distillation system for cleaning the sump solvent; this capability was not
a part of this evaluation. The LEVD also affords greater production flexibility, when pollution
prevention is an objective, because, unlike a conventional degreaser, there are no significant idling
losses between loads or downtime losses during shutdown.
On the national level, the total U.S. solvent demand is approximately 160 billion gal/yr
(Basta and Gilges, 1991). Approximately half this volume currently is recycled (on site or off site).
There is still a large potential for recycling more. A shift from offsite to onsite recycling would
prevent the transport of hazardous solvent wastes.
The largest single use for solvents in the United States is for vapor degreasing. As
much as 90% of this solvent is lost through air emissions. Other leading uses are dry cleaning (for
clothing) and cold cleaning (or immersion cleaning of soiled parts). The enclosed vapor degreaser
has potential for significant!^ reducing solvent use by preventing losses through air emissions.
28
-------
SECTION 4 i
ECONOMIC EVALUATION \
The economic evaluation compares the costs of each new technology to conventional
practice.
ATMOSPHERIC BATCH UNIT 1
The major operating costs associated with the disposal option and the atmospheric
batch unit are listed in Table 9. The disposal option operating costs total $16,529/yr and include
the following: j
• Annual purchase of 800 gal of virgin solvent at $10.50/galr based on plant records
• Annual disposal of 900 gal of spent solvent waste at $400/drum, based on plant
records i
• Seventeen 55-gal steel drums at $40 each
• Annual disposal labor estimate of 8 hr at $8/hr.
The operating costs for recycling with the atmospheric unit total $6,518/yr and include the
following:
• Virgin solvent purchase of 245 gal/yr at $10.50/gal to make up for processing
losses (during normal solvent use at the plant) and recycling losses (solvent lost to
distillation residue), based on Table 7 [
• Recycling operation labor at $8/hr based on 1 hr of operator involvement per batch
and 17 batches per year j
• Routine maintenance suggested by the vendor include (1) lid gasket replacement
once a year ($86/gasket); (2) checking the condenser for deposits and cleaning if
required once a year; and (3) periodic (monthly in this case) checks of the flow
switch to ensure that heating shuts off when water flow jto the condenser stops.
Estimated labor for the above maintenance is 12 hr/yr ;
• Energy costs based on 6.2-kW heater operating 12 hr/batch at $0.04/kWh
29
-------
TABLE 9. MAJOR OPERATING COSTS FOR ATMOSPHERIC UNIT
Item
Disposal Option
Virgin solvent
Disposal
- labor
- drums
— disposal fee
Atmospheric Unit
Virgin solvent
Operating labor
Routine maintenance
- spare parts
- labor
Energy
Cooling water
Disposal
- labor
- drums
— residue disposal
— still bags
Annual Usage
880 gal
8hr
17
900 gal
$245 gal
17 hr
1
12hr
1,265 kWh
18,360 gal
3
5
262 gal
17
Unit Cost
$10.50/gal
$8/hr
$40/drum
$400/55 gal
Total
$10.50/gal
$8/hr
$86/each
$8/hr
$0.04/kWh
$1/1,000 gal
$8/hr
$40/drum
$675/55 gal
$84/1 2 bags
Total
Annual Cost
$ 9,240
$ 64
$ 680
$ 6,545
$16,529
$ 2,573
$ 136
$ 86
$ 96
$ 51
$ 18
$ 24
$ 200
$ 3,21 5
$ 119
$ 6,518
• Water supply costs based on 1.5 gal/min to condenser for 12 hr/batch at
$1/1,000 gal
• Distillation residue disposal costs based on 262 gal/yr of residue (Table 7) at
$675/55-gal drum; still bags at $84/doz; 5 drums at $40/drum; and an estimated
disposal labor requirement of 3 hr/yr at $8/hr.
30
-------
Recycling, therefore, results in savings of $10,011/yr. !
The purchase price of the atmospheric batch unit is $12J99I5. A simple payback
period can be calculated as follows:
payback, years
purchase price
savings in annual operating costs
(1)
This result is a payback period of less than 2 years. This simple calculation does not include
factors such as taxes, inflation, installation, etc. A more detailed payback period calculation based
on worksheets provided in the Facility Pollution Prevention Guide (EPA,j 1992) is given in Appen-
dix F. The worksheets calculate the return on investment (ROD each yesir after the purchase. The
payback period (i.e., the year when the ROI exceeds 15%, which is the cost of capital) by this
i
calculation is less than 2 years. j
VACUUM UNIT !
The major operating costs of the vacuum unit are listed in Table 10. Also included in
this table are the costs associated with the disposal option. The disposal option operating costs
total $20,538/yr and include the following: !
• Annual purchase of 3,000 gal of virgin solvent at $3.57/gal, based on plant records
i
• Annual disposal of 3,000 gal of spent solvent waste at J?2.50/gal, based on plant
records !
• 55 steel drums costing $40/drum
• Annual disposal labor estimate of 28 hrs at $8/hr.
The operating costs for recycling total $2,255/yr and include the following!:
• Virgin solvent purchase of 246 gal/yr at $3.57/gal to make up for processing losses
(during normal solvent use at the plant) and recycling) losses (solvent lost to
distillation residue and air emissions) based on Table 2 I
• Recycling operation labor at $8/hr based on 1 hr of operator involvement per batch
and 55 batches per year j
31
-------
TABLE 10. MAJOR OPERATING COSTS FOR VACUUM UNIT
Item
Disposal Option
"Virgin" solvent
Disposal
- labor
- drums
- disposal fee
Recycling
"Virgin" solvent
Operating labor
Energy
Disposal
- still bottoms
- used oil0
- labor
- drums
Routine Maintenance
-oil0
- spare parts
- labor
Annual Usage
3,000 gal
16 hrs
55
3,000 gal
246 gal
55 hrs
985 kWh
136 gal
4 quarts
3 hrs
3
4 quarts
_b
16 hrs
Unit Cost
$3.57/gal
$8/hr
$40
$2.50/gal
Total
3.57/gal
$8/hr
$0.04/kWh
$2.50
$3.00/quart
$8/hr
$40/drum
$3.50/quart
$480 (max)b
$8/hr
Total
Annual Cost
$10,710
$ 128
$ 2,200
$ 7,500
$20,538
$ 878
$ 220
$ 39
$ 340
$ 12
$ 24
$ 120
$ 14
$ 480
$ 128
$ 2,255
0 The used oil wastestream can be minimized by changing the oil filter strictly in accordance
with the manufacturer's recommendations.
b The $480 cost for spare parts is a maximum, which assumes that the maximum number of parts
will be replaced as per the manufacturer's recommendations during each overhaul. Actual
maintenance costs could be lower. Appendix G summarizes the estimate.
32
-------
• Routine maintenance, as described in Appendix G i
i
e Energy costs based on the 2-hp vacuum pump operating 1J2 hr/batch at $0.04/kWh
• Distillation residue disposal costs based on 136 gal/yr of residue (Table 7) at
$2.50/gal; used oil disposal at $3/quart (can be minimized by changing the oil filter
regularly); 3 drums at $40/drum; and an estimated disposal labor requirement of 3
hr/yr at $8/hr. |
i
The savings due to recycling, therefore, amount to $18,283/yr.
The purchase price of the vacuum unit is $23,500 for the explosion-proof model.
A simple payback period can be calculated as follows:
payback, years =
purchase price '
savings in annual operating coiits
(2)
_
This results in a payback period of less than 2 years. This simple calculation does not include
factors such as taxes, inflation, installation, etc. A more detailed payback period calculation based
I
on worksheets provided in the Facility Pollution Prevention Guide is given in Appendix H.
LEVD
The economic evaluation involved a comparison between the operating costs of the
low-emission vapor degreaser (LEVD) and a conventional open-top vapor degreaser with the same
production capacity as the LEVD. The production rate of the LEVD was jdetermined during testing
by measuring the stage times that make up a cleaning cycle for the various runs.
Table 11 describes the time measurements performed during testing. For the
degreasing, condensation, air recirculation, desorption, and adsorption stages, stage time was
preset and is not expected to vary much based on operational factors. Carbon heatup stage time
was not preset but remained fairly constant over varying operational factors. The only real
variability, as seen in Table 11, is in vapor-fill (chamber heatup) times, which varied proportionately
with the mass of the workload (steel parts). Most important are Runs 5 and 8, which started with
parts that were cold and had oil on them, a situation typical of normal operation. Despite the
widely varying workload masses, the two runs had identical carbon heatup times. For each of
Runs 5 and 8, if all the stage times except vapor-fill time are added up, a 32.5-min period is
obtained that is relatively constant for all runs. In addition, prior tests done by the vendor with an
empty basket showed a vapor-fill time of 7.5 min, which would be the minimum vapor-fill time for
any workload.
33
-------
(O
z
cc
UJ
cc
o
u.
z
Q
Q
ixi
UJ
dn
C/3
UJ
I-
cn
UJ
LU
UJ
CO
ill
°
1^
uojjdjospv
(lasajd)
uoijdjosaQ
eaij uoqjBQ
inojpay j;v
(lasajd)
p.Bsuapuoo
(iasajd)
6uisBaj6aQ
INI JOdBA
•*- CO
o t;
+j CO •—
I£I
z%
6
c
cc
10 LO
, O5 Is* O O O5 O
1 CO CO . LO «*
fc ro
>f Q
ro -a
Q c
^ 0
S2 t— CN co QjLocDcocn
LI (f)
CO
w
.j ^
CN ro
"= °
o ^_
E "?
CO ^
-Q ro
o •*-•
& E
? 1
C H-
I S
1 CD Q3
o 1
3 O
0 |
o3
([^ T3
I I
i J
"co co
E ^ ,<
o o -o co
>• "g ro S
^ . '5 w 2
r- 4-J CT c O
TO <« Q> 3 •>
jc S CC a>
^_i *^ QJ 1—
lilli
•^ = 1 1 s
•*-• TD f- fO __
Tir *"• d) D)
0) ^ O P .E
— • c irt a)
OT 'E S -Q
£ .2 ro c c
> " .2 £ 2
CD ^ € | g
~ £ to w ^
CO CD *• CD 4.1
flfll
1^1-1
:= ~ S ^ ®
*~ m m CD
!U JO 0
34
-------
total
Run 5, with 165 Ib, had a vapor-fill time of 18 min and a
Run 8, with 915 Ib, had a vapor-fill time of 36.5 min and a total cyi
relationship between time and mass for steel parts is plotted in Figure
Runs 5 and 8. By intrapolating between these two runs, the workload
time of 60 min is determined to be 560 Ib. Note that when the workload
total cycle time is approximately 40 min (32.5-min fixed component
fill time) to 45 min.
cycle time of 50.5 min;
cle time of 69 min. This
5, based on the data for
mass that gives a total cycle
is 0.0 Ib (empty basket),
7.5-min minimum vapor-
plus
75
70
.5
£ KK
c DO
oT
P
•2 en
$
1
12 **
50
45
0
n^
2
^
00
4C
^
10
^
6
^
30
^
8(
^~
)0
^
-G
10
00
12
00
Mass of Parts Cleaned Per Cycle, Ib
Figure 5. Variation of LEVD cycle time for stee .
Therefore, the conventional degreaser selected for comparison has a production rate of
560 Ib of steel parts per hour. Based on information obtained from several vendors, Appendix E
describes the features of a typical vapor degreaser that can be expected to have a similar
production rate.
The major operating costs of the LEVD are listed in Table
12. Also included in this
table are the operating costs for a conventional open-top vapor degreaser with a similar production
capacity, as described in Appendix E. Vendor-quoted purchase price for the LEVD was $210,000.
35
-------
TABLE 12. OPERATING COSTS FOR VAPOR DECREASING
Item
Conventional Degreaser
Operating labor
Electricity
Cooling water
Maintenance
— labor
- materials
Net solvent loss
LEVD
Operating labor
Electricity
Maintenance
- labor
— materials
Annual Volume
$4,000 hr
25,500 kWh
480,000 gal
22 hr
-
2,642 Ib
333 hr
93,725 kWh
262.5 hr
_
Unit Cost
$8/hr
$0.04/kWh
$1/1,000 gal
$8/hr
-
$0.71 /Ib
Total
$8/hr
$0.04/kWh
$8/hr
Total
Total Cost
$32,000
$ 1,020
$ 480
$ 176
$ 88
$ 1,876
$35,640
$2,664
$ 3,749
$ 2,100
$ 2,100
$10,613
The annual operating cost of the conventional degreaser totals $35,640 and includes
the following: *
• Operating labor is the actual operator involvement in the process. Compared to the
LEVD, a conventional machine having the same production capacity is much smaller
in size and runs much smaller batches. Although the LEVD runs 560 Ib/batch/hr,
the conventional degreaser can attain this rate only by running several small batches
(cycles) per hour. This means that, over a period of 1 hour, operator involvement is
almost continuous, especially if each workload has to be raised, allowed to drain,
and then pulled out. Annual operator involvement for the conventional machine
therefore is 16 hr/day, 5 days/week, and 50 weeks/yr at $8/hr.
• Electric heat required is 6 kW, 17 hours/day (16 hr operation and 1 hr startup),
5 days/week, and 50 weeks/yr at $0.04/kWh.
• Primary cooling coil water consumption is 2 gal/min at $1/1,000 gal.
36
-------
• Annual maintenance labor cost was estimated as 2% of the unit's purchase price,
and maintenance materials cost was estimated as 1 % of the purchase price. The
capital cost of the conventional unit was estimated at §8,500,, based on information
from vendors. j
• Net solvent loss cost is the additional solvent that has to be purchased to make up
for air emissions. The net annual solvent loss is the difference in annual emissions
of the conventional unit (approximately 2,646 Ib/yr per Appendix E) and the LEVD
(approximately 4 Ib/yr per Section 3). Solvent price was based on a current market
price of $0.71/lb of perchloroethylene (PCE). :
The annual operating cost of the LEVD is $10,613 and includes the following:
Operator time of 5 min/cycle. Operator involvement is required only for unloading
and loading a basket of parts at the end of each cycle (batch). At 1 cycle/hr (for
560 Ib of steel parts), operating labor was based on 16 cycles/day, 5 days/week,
and 50 weeks/yr. j
i
Electricity costs were based on the energy requirement of 93,725 kWh/yr as
described in Appendix I.
Annual maintenance labor costs were based on 1 % of the LEVD purchase price and
maintenance materials costs were based on 1 % of purchase price. The purchase
price of the LEVD is quoted by the vendor as $210,000. j
The LEVD thus results in a savings in annual total operating costs of $25,027. The savings are
mainly due to reduced labor costs (due to larger batch size) and lower solvent requirement (due to
solvent recovery). i
A simple payback period can be calculated according to the following formula:
payback, years
purchase price
savings in annual operating costs
(3)
With a purchase price of $210,000 and an annual savings of $25,027 the LEVD pays for itself in
approximately 9 years. A more detailed economic analysis based on the worksheets in the Facility
Pollution Prevention Guide (EPA, 1992), is given in Appendix J. Although the above is a straight-
forward cost comparison between the LEVD and a conventional vapor degreaser of similar
production capacity, some other cost-benefit factors must be taken into account while making
economic decisions. One major consideration is that the LEVD does not require much of the
auxiliary equipment that may be required for a standard conventional vapor degreaser, if the user is
aiming to reduce workplace emissions to meet or anticipate increasingly stringent environmental
and worker safety regulations. Additional control devices for standard conventional degreasers
(e.g., increased freeboard ratio, refrigerated coils, lip exhausts, room ventilation) would add
37
-------
considerably to capital and operating costs. In contrast, the LEVD is a self-contained unit that
would require no additional facility modifications to achieve significant emission reduction.
Another consideration is the production rate of the LEVD. The above calculation was
performed for a production rate of 560 Ib/hr of steel parts (workload) because it is standard
practice with most vendors of conventional degreasers to quote capacities based on steel parts.
However, production capacity on the same machine can vary depending on the metal processed.
For example, many vendors of conventional equipment quote half the capacity of steel for similarly
shaped aluminum parts. This is ostensibly because the density of aluminum is half that of steel.
Thus, for the same basket or chamber volume, only half as many similarly shaped parts can be
processed per batch.
For the LEVD, variation in cycle time due to various workload metals can be estimated
based on the times measured for steel (Figure 5). The governing factor is the thermal diffusivity of
the metal. Thermal diffusivity (a) is the factor that ties in the heat conductivity (k), specific heat
(cp), and density (d) of the metal by the following relationship:
k
a =
d-cr
(4)
For the same shape and size of parts, the times required for different metals to reach vapor
temperature are inversely related to their respective thermal diffusivities.
* tmetal ~ °steel * tsteel
(5)
Therefore, Appendix K lists the k, cp, and d values for steel, aluminum, brass, and copper.
Appendix K also calculates the total cycle times for the various metals in the LEVD. Note that,
regardless of the metal or workload mass, the total cycle time has a fixed component of 32.5 min
(discussed earlier in this section) and a minimum vapor-fill time of 7.5 min. Only the variable
component, the vapor-fill stage time, is affected by the metal and mass. Based on Appendix K, the
total cycle times for various metals versus production rates are plotted in Figure 6. Thus, brass
and copper can be processed faster than steel in the LEVD. Aluminum can be processed faster up
to a point. This point is determined, for a certain shape of parts, by the maximum mass of
aluminum parts that can fit into the basket and be processed per cycle. For the 915 Ib of steel
parts processed in Run 8 of this testing, the basket was nearly full. The vendor literature lists the
maximum capacity of this LEVD model as 1,100 Ib of steel parts/cycle. A full batch (basketful) of
aluminum parts (of the same shape and size as steel) would weigh only 380 Ib. Above 380 Ib, the
parts would have to be processed in the next cycle. Hence, the production rate for aluminum is
restricted to 44.5-min cycles containing 380 Ib/cycle. In Figure 6, the line for aluminum rises
38
-------
100
30
200
400 600 BOO
Mass of Parts Cleaned Per Cycle, Ib
1000
1200
Figure 6. Variation of LEVD cycle time for various metals.
sharply (by 32.5 plus 7.5 min) at 380 Ib to 84.5 min and continues with the same slope. With
44.5-min cycles, the average processing rate over a 16-hr day for aluminum on this machine is
497 Ib/hr. Brass and copper, however, can be processed faster than steel. The closer to maximum
load that a user can run in each cycle, the better is the time efficiency, j
i
Another factor that may affect cycle time is the shape of 1the parts. Parts that have
recesses where solvent could get trapped should be arranged in the basket in such a way that the
solvent liquid drains out rather than stays in. Alternatively, other features offered by the vendor,
i
such as oscillating or rotating baskets, should be used. If this is not dqne, either the air recircula-
tion stage time will have to be increased, or the unit will go into a loop of several adsorption cycles
until the chamber concentration falls below 1 g/m3. j
39
-------
SECTION 5
QUALITY ASSURANCE
A Quality Assurance Project Plan (Battelle, 1992b) was prepared and reviewed by EPA
before testing began. This QAPP contains a detailed design for conducting this study. The
experimental design, field testing procedures, and laboratory analytical procedures are described.
The QA objectives outlined in the QAPP are discussed below.
Onsite TESTING
Onsite testing was conducted as planned, except for the following. The plan was to
evaluate the vacuum unit on both methylene chloride (MC) and methyl ethyl ketone (MEK).
However, on reaching the site, the contact at the plant discovered that, a few weeks earlier, shop
floor workers had already substituted the MEK used in the marking operation with a less hazardous
solvent. Hence, the vacuum unit could be tested only on MC.
For the low-emission vapor degreasing (LEVD) unit, it should be noted that the
extended desorption cycle required at the end of the day to regenerate the carbon had been
eliminated by the manufacturer by the time of the testing.In the new design, a cooling coil has
been added at the bottom of the working chamber to condense out most of the solvent before it
gets to the carbon. The improved desorption stage during each normal cleaning cycle is sufficient
to regenerate the carbon.
For the LEVD testing, one of the efforts planned was isolation and volume measure-
ment of the condensed solvent recovered during a cycle, especially during the extended desorption
stage. The objective was a double verification that solvent prevented from ambient release during
a cycle (as indicated by the FID measurements) is captured on the carbon and recovered. With the
extended desorption cycle eliminated, this aspect of the testing could not be done quite as planned.
Piping was changed to collect the condensate from the newly added cooling coil. Condensate was
collected in glass containers packed in dry ice to avoid evaporative losses of the hot liquid. The
volume obtained was 1,500 mL during the condensation stage. Another 100 mL was collected
during the desorption stage (left over from the previous run, approximately equal in volume to
solvent adsorbed in the current run). The recovery of the solvent thus was verified. On the
previous day, 88 L (approximately 23 gal) of solvent had been filled into the sump. The volume of
40
-------
solvent remaining in the sump after completion of the run was not measured because the solvent in
the sump remains hot for several hours (overnight the solvent cools only to 70°C) due to the
enclosed design, and further testing had to be continued. j
The QAPP stated that intermittent ambient air measurements would be made outside
the unit with a FID probe. This was improved upon because of an added data-capture system on
the FID and because the relatively operator-free cycle of the LEVD made possible continuous
measurements all around the unit for several of the runs.
i
LABORATORY ANALYSIS \
Laboratory analysis was performed as planned. The ME:K acidity measurements
(ASTM D1613) reported in Table 5 were not mentioned in the QAPP but were conducted as part of
the purity analysis (ASTM D2804); there was some interpretive benefit in reporting them separate-
ly. Because of the interpretive success of the MEK purity analysis, a similar analysis (not
mentioned in the QAPP) was performed on a recycled sample and the "virgin" sample of MC. No
purity analysis was done on the spent MC sample because it was too viscous and contaminated to
be analyzed without extensive additional steps. Absorbance measurements (UV VIS) were
conducted on the solvent samples as another simple indicator of solvent quality.
i
Table 13 describes the detection limits and precision of thei duplicate analysis on the
I
solvent samples as per the QAPP. Except for the analysis of nonvojatiles on MEK, all other
parameters had acceptable precision, within the range specified in'thei QAPP. The analysis of
I
nonvolatiles on MEK exceeded the range by 1 % in sample values clo^se to the detection limit
i
(recycled sample). This small deviation should not have had a significant effect on the results
because the difference between spent and recycled values is several orders of magnitude greater
than the relative difference between the duplicates. For consistency, all duplicate analyses were
performed on the recycled sample for both MEK and MC. In most of thje preceding tables in this
report, results for both duplicates have been reported.
LIMITATIONS AND QUALIFICATIONS
The above QA data indicate that the results of the onsite and laboratory testing can be
considered a valid basis for drawing conclusions about product quaility, pollution prevention
i
potential, and economics. Site-specific economic and waste generation information for the
atmospheric and vacuum units was obtained from plant records going back to the year before each
new technology was installed. For the LEVD pollution prevention and economic evaluation,
41
-------
TABLE 13. PRECISION OF SOLVENT ANALYSIS
Parameter
Color
Nonvolatiles
Specific gravity
Water content
PH
Acid acceptance
Purity
Conductivity
Analytical
Method
ASTM D1209b
ASTM D2108C
ASTM D1353b
ASTM D21090
ASTM D891b
ASTM D21110
ASTM D1364b
ASTM 03401°
ASTM 02110°
ASTM 02942°
ASTM D2804b
EPA 120.1b
EPA 120.1°
Sample
Value8
5
5
2.6
20.37
0.827
1.286
5.42
0.25
6.7
0.004
85.02
3.30
137
Duplicate
Value8
5
5
2.0
17.88
0.821
1.287
5.56
0.24
6.6
0.005
85.54
3.40
136
Precision
(% RPD)
_d
_d
26.1
13.0
0.7
0.0
2.6
4.1
1.5
22
-0.6
-0.0
0.7
Detection
Limit
NAe
NA
1 mg/100 mL
1 ppm
0.001
0.001
0.01%
0.01%
0.1
0.001%
0.01 %
1 jwmho/cm
1 jumho/cm
8 Measurement units are given in "Detection Limit" column.
b MEK.
0 MC.
d Duplicates target same color standard number. Precision acceptable.
0 NA = Not acceptable.
baseline information for a conventional open-top vapor degreaser was taken from (1) information
provided by several vendors of conventional equipment contacted during the study, (2) a previous
study conducted for the WRITE Program (Battelle, 1992a), and (3) a previous EPA study (EPA,
1989) of several varieties of vapor degreasers.
For the vacuum unit, the product quality evaluation was somewhat affected by the
fact that the "virgin" solvent was actually a commercially recycled solvent. Therefore, the vacuum
unit was tested on an unusually severe matrix, in that existing volatile impurities in the purchased
"virgin" solvent had accumulated in the spent solvent over several months of use and repeated
on-site recycling. Better recycled MC purity test results probably could have been obtained if the
spent solvent had not contained volatile impurities from external sources. This underscores the
need for users of more than one solvent to segregate solvent wastes before onsite recycling.
42
-------
Repetitions of the runs on the atmospheric and vacuum
because of the low availability of spent solvent at the sites during testing.
for a single test run on each unit. However, it is to be noted that the sites
units for a long time without experiencing major difficulties. The test
considered by the sites to be typical of the past several runs in recycled
quality.
units were not possible
Hence, the results are
have been using these
runs at both sites were
solvent quantity and
43
-------
SECTION 6
CONCLUSIONS AND DISCUSSION
The three technologies evaluated in this study were (1) an atmospheric batch distilla-
tion unit, (2) a vacuum heat-pump distillation unit, and (3) a low-emission vapor degreaser. The
objective was to determine the suitability of each unit to its respective application, rather than to
compare among the three units. All three units performed satisfactorily during the testing and were
able to recover significant quantities of solvent. A large portion of the approximately 160 billion
gal/yr of solvent consumed annually in the United States could be recycled. Between offsite and
onsite recycling, onsite recycling is preferable because transportation hazards are reduced. The
largest single use for solvent in the United States is for vapor degreasing, followed by dry cleaning
and cold cleaning (immersion cleaning of parts). The LEVD has great potential for reducing solvent
loss in vapor degreasing through air emissions.
ATMOSPHERIC BATCH UNIT
The atmospheric unit greatly improved the quality of the spent methyl ethyl ketone
(MEK) solvent feed. The quality improvement was immediately apparent from the appearance and
colorlessness of the recycled solvent and was confirmed by the laboratory analysis. Standard
solvent quality-monitoring parameters such as specific gravity, absorbance, and conductivity
showed good agreement between recycled and virgin solvent values. Nonvolatile matter was
virtually removed from the spent solvent. MEK purity rose from 78% in the spent solvent to 85%
in the recycled solvent. The recycled value was lower than the >99% pure virgin grade.
However, because the impurities are believed to be mainly paint thinners (solvents), they should
not cause any problem when reused to clean spray painting lines between colors. The presence in
the recycled solvent of 5% water (possibly due to a small leak in the water-cooled condenser) is of
some concern, although the site did not report any adverse effect on the painting operation.
Compared to disposal, the atmospheric unit evaluated in this study has good solvent
waste reduction potential. At the test site, annual hazardous solvent waste generated for disposal
was reduced from 880 gal to 262 gal. A sidestream of 18,360.gal/yr of cooling water (tapwater)
is generated during recycling, but this should not be a problem because the water does not come
44
-------
into contact with any contamination and can be reused elsewhere as process water. The recycling
unit is easy to install and operate safely.
The economic benefits of the atmospheric unit include reduced operating costs and a
payback period of less than 2 years. One economic benefit that is difficult to quantify is the
reduction in potential liability as a result of eliminating the transport of large volumes of hazardous
solvent waste. - j -
i
i
VACUUM HEAT-PUMP UNIT |
The vacuum unit considerably improved the quality of the;spent methylene chloride
(MC) solvent. This was apparent from the visual appearance of the solvent and was confirmed by
the laboratory analysis. The evaluation was complicated by the fact that the "virgin" solvent was
actually a commercially recycled solvent. This "virgin" grade itself was analyzed to be 90% pure
MC. Hence, the spent solvent fed to the vacuum unit probably contained significant amounts of
volatile impurities accumulated over several previous recycles at the site. Even then, an 86% MC
purity was achieved by the vacuum unit; the 14% impurities are likely to have been other volatile
solvents. Nonvolatile matter was virtually removed. Standard solvent quality-monitoring parame-
ters, such as specific gravity, absorbance, and conductivity, showed good agreement between
recycled and "virgin" values, indicating that the recycled material was a;fairly close approximation
of the "virgin" grade. The pH of water extracts of the spent, recycled, ajnd virgin samples showed
that the recycled solvent had a pH close to 7 (the "virgin" sample pH), (compared to a potentially
corrosive spent sample pH value of 5. The corrosion test showed tfiat spent solvent caused
corrosion of steel, but that the recycled solvent caused no noticeable corrosion.
The vacuum unit evaluated in this study has good solvent i waste reduction potential
compared to disposal. At the test site, annual hazardous solvent waste generated for disposal was
reduced from 3,000 gal to 136 gal. At this test site, a largely avoidable siclestream of 218 gal/yr
of air emission also was generated during recycling but can be avoided; by operating the vacuum
pump at a slower rate. The recycling unit is easy to install and operate Seifely.
The economic benefits of the vacuum unit include reduced operating costs and a
payback period of less than 2 years. One economic benefit that is difficult to quantify is the
reduction in potential liability as a result of eliminating the transport of lairge volumes of hazardous
solvent waste. !
45
-------
LOW-EMISSION VAPOR DEGREASER
Several runs with well-oiled steel parts verified that the low-emission vapor degreaser
(LEVD) was performing its basic function. One aspect of the enclosed design that was noted was
its potential for reducing water contamination of the solvent. In the enclosed design, only a limited
amount of ambient air (equal to the free volume of the loaded chamber) comes into contact with
the cooling surfaces of the unit. Hence, there is lower atmospheric contribution of water to the
liquid solvent in the LEVD. Lower water contamination implies lower depletion of solvent stabilizers
and acid formation by hydrolysis.
The LEVD reduced air emissions by more than 99% compared to the estimated air
emissions from a typical conventional open-top vapor degreaser (560 Ib of steel parts/hr capacity,
with a 0.75 freeboard ratio, primary cooling coil, electric hoist, and no lip exhausts). It is true that
air emissions from conventional degreasers can be reduced by improved operating practices and by
modifications in the equipment (e.g., increased freeboard ratio, refrigerated coils, covered opening,
reduced room draft/lip exhaust velocities). However, it would be difficult for any conventional
degreaser to match the pollution prevention performance of the completely enclosed design used in
this study. The perimeter ambient survey indicated that the OSHA exposure limit of 25 ppm (8-hr
TWA) would be met by the LEVD. The pollution prevention potential of this unit is further
enhanced by its ability to perform as a liquid solvent distillation system for cleaning the sump
solvent; this capability was not a part of this evaluation. When pollution prevention is an objective,
the LEVD also affords greater production flexibility because, unlike a conventional degreaser, the
LEVD has no significant idling losses between loads or downtime losses during shutdown.
The economic benefits of the LEVD include reduced operating costs and a payback
period of approximately 10 years. If the objective is to meet current or anticipated stringent
environmental and occupational safety regulations, the LEVD is a self-contained piece of equipment
that will meet this objective.
46
-------
SECTION 7
REFERENCES
Basta, N. and K. Gilges. 1991. "Recycling Everything, Part 4: The Recycling Loop Closes for
Solvents." Chemical Engineering. June.
Battelle. 1992a. An Automated Aqueous Rotary Washer for the
U.S. Environmental Protection Agency Project Summary. EPA/600/SR-92
Battelle. 1992b. Quality Assurance Project Plan for Onsite Solvent Recovery. Columbus, Ohio.
Handbook. McGraw-Hill Book
Perry, R. H., and C. H. Chilton (Eds.). 1973. Chemical Engineers'
Company, New York, New York.
Surprenant, K. S. 1992 (draft). Study of the Emission Control
Freeboard on Open Top Vapor Degreasers. Prepared for Emission S
Division, Office of Air Quality Planning, U.S. Environmental Protection
Chemical Company.
Effectiveness of Increased
Standards and Engineering
Agency, by the Dow
U.S. Environmental Protection Agency. 1992. Facility Pollution Prevention Guide. EPA/600/R-92/-
088.
U.S. Environmental Protection Agency. 1989. Alternative Technology Control Documents —
Halogenated Solvent Cleaners.
47
Metal Finishing Industry.
/188.
-------
APPENDIX A
ADDITIONAL FEATURES OF THE THREE TECHNOLOGIES
The vendors of the three technologies described in Section 1 of this report offer
additional features to suit individual users. These features are discussed here.
ATMOSPHERIC DISTILLATION
In addition to the unit described in Section 1, other standard models of 5 and 15 gal
per batch capacity are available and larger units can be designed. Vacuum attachments can be
retrofitted on many of the models for higher boiling solvents. Air-cooled condensers are offered as
an option on some of the models. Another model incorporates a cooling system consisting of a
water/glycol reservoir with fan. This water/glycol cooling system is designed for solvents that boil
above 150°F and comes in a weather-proof enclosure. Potential applications include chlorinated
solvents, alcohols, aliphatic petroleum hydrocarbons, aromatics, esters, chlorofluorocarbons
(CFCs), ketones, terpenes, and new replacement solvents such as N-methyl-2-pyrrolidone (NMP),
etc.
VACUUM HEAT-PUMP DISTILLATION
The vendor offers six continuous models varying in distilling capacity. Lower capacity
models have a horizontal heat exchanger with one pump, and the higher capacity models have a
vertical heat exchanger with two pumps. The same heat-pump concept also is offered in batch
design. The batch heat-pump models are designed to further concentrate distillation residues from
the continuous models or other atmospheric batch distillation units. The performance claimed for
these second-stage stills is to concentrate feed material containing 50% solvent to still bottoms
with 5% solvent.
LOW-EMISSION VAPOR DEGREASER
The unit described in Section 1 is intended mainly for medium to large parts. The
vendor offers larger models with capacities of 2,200 Ib and 3,300 Ib of steel parts (or higher) per
48
-------
cleaning cycle (or per batch). Features such as rotating baskets or basket oscillators are available
for cleaning small parts. Other optional features include cold dip, hot immersion, ultrasonic
cleaning, side tray loading of parts, and conveyorized loading. Future modifications are aimed at
cutting down the cycle' time by eliminating the carbon heatup and desorption stages through the
design of twin carbon beds, so that one bed can be desorbed while the other is adsorbing.
49
-------
APPENDIX B
ANALYTICAL TESTING OF SPENT, RECYCLED, AND VIRGIN SOLVENTS
The samples of spent, recycled, and virgin samples collected during the evaluation of
the batch distillation and vacuum units were analyzed by standard ASTM methods (and one EPA
method). The objective was to compare the quality of the recycled and virgin solvents on the one
hand, and of the recycled and spent solvents on the other. The recycled versus virgin comparison
indicates how closely the recycled sample resembles virgin grade; the recycled versus spent
comparison indicates the improvement in quality brought about by recycling. The following
parameters were used as the basis for this comparison. Table B-1 lists the titles of the standard
methods used in this study. In the following description, it should be noted that the same
parameter may be measured by slightly different methods depending on whether the solvent is a
volatile organic solvent (e.g., methyl ethyl ketone [MEK]) or a halogenated organic solvent (e.g.
methylene chloride [MC]).
Appearance: Visual appearance of a solvent often is used as an initial indicator of
quality. Because most industrial solvents are transparent, any suspended or floating matter,
sediment, turbidity, or free water is easily seen. ASTM D3741 covers the procedure for this visual
determination.
Color: Color often is one of the first indicators of contamination of essentially clear
organic solvents. The greater the color, the greater is the potential for interference with solvent
functions such as cleaning and degreasing. ASTM D1209 (for clear organic liquids) and ASTM
D2108 (for halogenated organic solvents) cover the visual measurement of color on a standard
platinum-cobalt scale.
Nonvolatile Matter: Volatile solvents are used in paint, varnish, lacquer, and other
related products in which the presence of residue may compromise quality. Residue also can affect
the cleaning effectiveness of halogenated solvents in degreasing operations. The ability of the
recycling processes to remove such residue from spent solvent is tested by ASTM D1353 (for
volatile solvents) and ASTM D2109 (for halogenated organic solvents).
50
-------
TABLE B-1. STANDARD TEST METHODS FOR SOLVENT ANALYSIS
Test Designation
Title
ASTM D3741-85 (91)
ASTM D1209-84 (88)
ASTM D2108-85
ASTM D1353-90
ASTM D2109-85
ASTM D891-89
ASTM D2111-85
ASTM D1364-90
ASTM D3401-85
ASTM D2110-85
ASTM D2942-86
ASTM D2251-85
ASTM D2804-88
EPA 120.1
ASTM D1613-91
EPA Method 110.3
Appearance of Admixtures Containing Halogenated Organic Solvents
Color of Clear Liquids (Platinum-Cobalt Scale)
i
Color of Halogenated Organic Solvents and their Admixtures
(Platinum-Cobalt Scale) j
Nonvolatile Matter in Volatile Solvents for Use in Paint, Varnish,
Lacquer, and Related Products [
Nonvolatile Matter in Halogenated Organic Solvents and their
Admixtures
i
Specific Gravity, Apparent, of Liquid Industrial Chemicals
Specific Gravity of Halogenated Organic Solvents and their Admixtures
Water in Volatile Solvents (Fischer Reagent Titratiori Method)
Water in Halogenated Organic Solvents and their Admixtures
pH of Water Extractions of Halogenated Organic Solvents and their
Admixtures
Total Acid Acceptance of Halogenated Organic Solvents (Nonreflux
Methods)
Metal Corrosion by Halogenated Organic Solvents and their Admixtures
Purity of Methyl Ethyl Ketone Using Gas Chromatography
Conductance i
Acidity in Volatile Solvents and Chemical Intermediates Used in Paint,
Varnish, Lacquer, and Related Products j
s
Absorbance !
Specific Gravity: The specific gravity of each organic solvent is used in industry to
determine the identity and purity of the solvent being used. Specific gravity change also is
measured by users to monitor deterioration of solvent performance and cleanliness. This parameter
is a simple hydrometer measurement that is described in ASTM D891 (for liquid chemicals in
i
general) and in ASTM D2111 (for halogenated solvents in particular). j
i
Water Content: Solvents are used under a variety of conditions that may be adversely
affected by the presence of water. Spent solvent may contain water du« to contamination during
operation or simply due to atmospheric condensation. The ability of the recycling process to
51
-------
eliminate water from the spent solvent will be measured by ASTM D1364 (for volatile solvents) and
ASTM D3401 (for halogenated organic solvents).
pH: Halogenated solvents with alkaline stabilizers (amine-types) generally have a pH
ranging from 7 to 11. Virgin and recycled solvents should fall within these ranges. Halogenated
solvents are prone to a lowering of these values during use due to the formation of acids (e.g.,
HCI). An acidic environment is more corrosive. In ASTM D2110, the solvent sample is shaken
with distilled water and the pH of the water extract is measured.
,j-*
Total Acid Acceptance: The tendency of halogenated solvents to form acidic
conditions is curtailed by adding stabilizers as described above. The question during recycling by
distillation is whether the stabilizers are distilled with the solvent or decompose. ASTM D2942
measures the total acid acceptance properties of virgin and recycled solvents for both alkaline and
neutral stabilizers.
Metal Corrosion; Corrosion is an important consideration for halogenated solvents
used in degreasing and cleaning metal parts. This test (ASTM D2251) often is used as a guide in
selecting or eliminating solvents used for degreasing or cleaning. For this test, polished strips of
metal (steel and aluminum were used in this study) are immersed in the solvent and heated at
reflux for 60 minutes. Similar strips also are kept immersed in solvent in closed containers for 10
days. At the end of this period, the strips are removed and examined visually for evidence of
tarnish or corrosion.
Purity: ASTM D2804 describes the determination of the purity of MEK by gas
chromatography. This method provides a quantitative indication of impurities. Although not
planned in the QAPP, a similar analysis was performed on the MC recycled and "virgin" samples.
Conductivity: Liquid conductivity, as determined by Standard Method EPA 120.1 for
industrial wastes, often is measured by solvent users as a quick test for contaminants. It provides
a measure of any dissolved ionic impurities.
Acidity: This parameter was not planned in the QAPP as a stand-alone analysis but
was part of the purity determination. However, it is reported separately in the report for its
interpretive value. This parameter (ASTM D1613-91) measures total acidity as acetic acid. Acidity
may be present in the spent solvent as a result of contamination during use or decomposition
during storage.
52
-------
Absorbance: The absorbance of the solvent in the 400 to (500 nm wavelength range
was measured as a quick check on the degree of contamination. Contamination in solvent affects
the transmittivity of light though the solvent. The UV-VIS absorbance meter was zeroed out using
clear reagent-grade solvent (MEK or MC), and the test samples were measured with respect to this
baseline. This test is based on EPA Method 110.3.
53
-------
APPEMDIXC
CALIBRATION OF FID INSTRUMENTS
The two flame ionization detectors (FIDs) used in the testing — the "chamber FID" and
the "ambient FID" — were calibrated with propane as well as perchloroethylene(PCE)-in-air
standards. The PCE standards were verified with a second standard obtained from a different
source. Instruments were zero-adjusted with a zero gas (hydrocarbon-free air). The main objective
of the chamber FID measurements was to ensure that chamber concentration at the end of each
cycle (when the lid is opened) was <150 ppm. The main objective of the ambient FID measure-
ments was to ensure that at locations close to the LEVD, readings did not rise significantly above
the facility ambient levels. The procedures used were as follows.
At the beginning of each test day, the instruments were first zeroed out with the zero
gas. The instrument was then calibrated to three propane standards — 945.5 ppm, 94.28 ppm,
and 9.616 ppm.
On Day 1 the following sequence of PCE standards was run to adjust the response:
Initial
ppm of Chamber FID
PCE Standard Readout
Beginning of Day 1 :
94.28 (propane recheck)
733
zero gas (recheck)
9.85
zero gas (recheck)
95.7
733
Noon;
95.7
End of Dav:
9.85
95
870
0.01
6.9
0.01
97.0
985
95.5
9.83
Initial
Ambient FID
Readout
95
865
0.01
6.9
0.02
95.0
1,000
95.5
9.82
Final
- Readout Adjusted to -
Chamber FID Ambient FID
NAa
732
NA
9.85
NA
95.7
NA
NA
NA
NA
732
NA
9.85
NA
95.7
NA
NA
NA
a No adjustment performed.
54
-------
As seen in the preceding table, the instrument was first rechecked with propane. Then the
733 ppm PCE standard was run, and the readouts showed 870 and 860 ppm. The readout was
i
adjusted (with adjust knob) to read 732 and 732 ppm. This was a convenient, direct way of
reading the instrument for PCE rather than leaving it at the propane adjustment and having to
multiply each reading with a relative response factor. i
Zero gas was run again as a check. The 9.85 ppm PCE standard was run next. The
FIDs initially read 6.9 and 6.9 ppm. FIDs were then adjusted to read 9.85 and 9.85. After another
zero check, the 95.7 PCE was run. The instrument read 97 and 95 ppm.I Both FIDs were adjusted
to read 95.7. All adjustments (with the adjust knob) were stopped at this point because maximum
accuracy was required in the 0 to 100 ppm range. This is because the chamber concentration (at
the end of the cleaning cycle) was expected to be in the higher part of this range and because
ambient concentrations were expected to be in the lower part of this range.
The 733 ppm standard was run to verify linearity at higher levels. The FID readings were
985 and 1,000 ppm. This indicated that, at very high levels, some linearity had been lost. No attempt
was made to readjust the instruments, in order to leave the accurate settingis at 95.7 and 9.85 intact.
At noon, a single-point check was run with 95.7 ppm PCE standard and both FIDs were
found to be holding well. At the end of the day, a check was run on the 9.85 ppm range, and both
i
FIDs again were found to be holding well.
On the second day, calibration was repeated, first for instrument calibration with
945.5, 94.28, and 9.616 ppm propane standards, then with the PCE standards.
Initial Initial Final
ppm of Chamber FID Ambient FID - Readout Adjusted to -
PCE Standard Readout Readout Chamber FID Ambient FID
Beginning of Day 2:
9.85 9.1 8.4 N
733 745 780 r<
Noon:
733 741 740 IN
\a NA
JA NA
IA NA
zero gas (recheck) 0.01 0.01 NA NA
End
9.85 10.1 10.0
of Day:
9.85 9.85
733 725 862 NA NA
No adjustment performed.
55
-------
I
On the second day, the FIDs were calibrated mainly at 9.85 ppm. The 733 ppm standard was run
to check linearity at this high level. On this day, response appeared to be somewhat more linear at
the higher level also. At noon, a calibration check was conducted at 733 ppm and 9.85 ppm PCE,
and the FIDs were found to be holding fairly well. The end-of-the-day check was run on 733 ppm
PCE. In general, the objective was for the calibration checks to be within ±15% of the standard.
For the end-of-the-day check on 733 ppm standard, however, the ambient FID readout showed
862 ppm (17.6% deviation from standard). This slight deviation is not expected to affect the
results significantly.
56
-------
APPENDIX D
LEVD CHAMBER CONCENTRATIONS AT END OF CYCLE
For Runs 1, 3, 5, and 6 of this testing, the chamber concentrations at the end of the
cleaning cycle were tracked with a flame ionization detector (FID) probe. The resulting data were
directed to a computer data capture system to generate the following charts. Normally, a
photoionization detector (PID) monitor on the unit checks the concentraticin in the chamber 60 sec
into the adsorption stage. If the concentration is less than 1 g/m3 (150 ppm of PCE), the lid opens
and the cycle ends. If not, the unit continues the adsorption stage up to 240 sec and rechecks the
concentration. This is repeated until the desired concentration is achieved. However, for this
testing the unit was programmed to run the full 240-sec adsorption stage and release the lid.
57
-------
o
o
o
-o
CO
o
o
IO
o
o
O
o
111
CO,
LU
o
-o
CM
O
O
c
o
0>
u
o
O
•*••
I
o
to
<5
JD
10
U
O)
_c
ra
_«
O
I
CO
CO
00
10
°<3
o
u
HI
w
h.
Q
a
£
3N31AHiaoyOlHOH3d SV
NOIIVyiN3ONOO
58
-------
c
o
0)
u
o
O
•*••
-------
.2
"-P
i
Q)
U
O
U
•+•*
0>
"o
<5
.O
(0
6
O)
_c
(0
_0)
U
in
*fc
i
CO
00
(0
_0)
o
o
U
Q
O)
(1H02O) 3N31AH13OHO1HOy3d SV IMdd NOI1VH1N3ONOO
60
-------
o
0)
u
o
U
*-*
w
~5
in
E
ra
O
O)
(0
w
O
3
CC
"55
OQ
eo
m
oo
(0
_OJ
u
o
o
UJ
t
Q
i
Q
o>
3N31AH13OdOlHOd3d SV lAIdd NOIlVdlNSONOO
61
-------
APPENDIX E
FEATURES OF THE CONVENTIONAL DEGREASER USED AS A COMPARISON
The baseline for comparing the pollution prevention and economic benefits of the
LEVD was a conventional open-top vapor degreaser. The conventional unit selected was the size
yields approximately the same production rate as the LEVD. The LEVD can process up to 560 Ib/hr
of steel parts (see Section 4 of this report). Based on information obtained from several vendors,
such a conventional unit typically would have an opening of 4.5 ft2. A standard unit comes with
primary cooling coils and a freeboard ratio of 0.75. The heating requirement is 6 kW and water
flow though the primary condenser is 2 gal/rnin. It should be noted that the physical dimensions of
the vapor chamber of the selected conventional unit are much smaller than the LEVD. The same
processing rate (560 Ib/hr) is maintained by running smaller, more frequent batches.
Emissions from conventional vapor degreasers have been estimated in previous EPA-
sponsored studies. In a previous 1992 study (Battelle, 1992a)*, a conventional open-top
degreaser was found to be losing 6,145 Ib/yr from a total feed of 6,190 Ib/yr (over 99% of feed)
due to air emissions. Another study by the Dow Chemical Co. (Suprenant, 1992) found that open-
top vapor degreasers with 0.5 and 0.75 freeboard ratios were emitting 0,373 and 0.273 Ib/ft2/hr
of solvent.
A more detailed EPA-sponsored study (EPA, 1989)* had been conducted previously to
estimate emission rates from conventional degreasers with varying control options. Losses due to
air emissions were estimated for three categories — workload losses, idling losses, and downtime
losses. Emissions, especially workload losses, varied depending on equipment- and operation-
related factors. Workload and idling losses, measured in terms of Ib of solvent per ft2 of opening
per hr (Ib/ft2/hr), generally were independent of the type of solvent used. Downtime losses were
dependent on the type of solvent (vapor pressure). Therefore, during a 24-hr period, air emissions
depended on the time distribution between working (workload processing), idling, arid downtime for
the degreaser.
For this study, the comparison was based on a degreaser that is operated (working)
continuously for 16 hr (no idling period) and shutdown (downtime) for 8 hr. In the EPA (1989)
study, 25 tests were conducted on various degreasers to estimate workload losses. Workload-
" References are listed in Section 7 of this report.
62
-------
related losses include losses due to disturbance of the air-vapor boundary during entry-exit of the
workload, processing of the workload, and liquid dragout on the workload., All the tests were
performed using electric hoists to move the workload. None of the degreasers had a lip exhaust
system. The tests included a wide range of room ventilation conditions. For the 25 tests, the
(•
average emission was 0.171 Ib/ft2/hr with a standard deviation of 0.141 Ib/ft2/hr. These tests
were conducted on various sizes of degreasers. Of the 25 tests, 12 tests were conducted on
degreasers with 4.5-ft2 openings. For these tests, the average and standard deviation were 0.133
and 0.046 Ib/ft2/hr, respectively. Of these 12 tests, 6 were conducted on degreasers with a
freeboard ratio of 0.75. For these six tests, the average emission and'standard deviation were
0.147 and 0.051 Ib/ft2/hr, respectively. This workload-related average emission of 0.147 Ib/ft2/hr
was used as the baseline in this study for comparison with the LEVD. j
[Downtime losses, which were expected to be an order of magnitude below workload
losses, were assumed to be minimal; these losses can be minimized by covering the vapor
degreaser during standby. Idling losses in the EPA (1989) study were ifourid to average 0.105,
which is the same order of magnitude as the working losses. Therefore, idling losses were not
quantified separately from working losses.]
The conventional open-top vapor degreaser selected for comparison with the LEVD,
therefore, has the following characteristics: !
4.5-ft2 opening
Approximately 560 Ib/hr production rate for steel parts
0.75 freeboard ratio
primary cooling coils
6-kW heater
2 gal/min tap water flow through primary cooling coils
Working emissions of 0.147 Ib/ft2/hr.
Based on information supplied by various vendors, such a degreaser can be expected to cost
I
approximately $8,500 (purchase price). I
63
-------
APPENDIX F
ECONOMICS OF ATMOSPHERIC UNIT
The costs utilized in the economic spreadsheets for the atmospheric unit (Tables F-1 to
F-4) are described below:
• The cost of equipment of $14,300 includes the listed price of $12,995 plus 10% to
cover freight, taxes, insurance, and spare parts.
• Materials, installation, plant engineering, and contingency costs are estimates based
on experience.
• Working capital is based on a 1-month supply of materials (still bags and drums).
Startup costs are estimated.
• Information on debt, depreciation, taxes, inflation, and cost of capital are general
assumptions.
• Differences in operating costs are based on. costs listed ,in Table 9 of the main
report. Supervision and overhead costs are general assumptions.
The payback period can be determined from Table F-4 as the year in which the return on invest-
ment (ROD exceeds 15% because 15% is the cost of capital (Table F-1).
64
-------
TABLE F-1 . CAPITAL COST OF THE ATMOSPHERIC UNIT
INPUT
CAPITAL COST
Capital Cost
Equipment
Materials
Installation
Plant Engineering
Contractor/Engineering
Permitting Costs
Contingency
Working Capital
Startup Costs
% Equity
% Debt
Interest Rate on Debt, %
Debt Repayment, years
Depreciation Period
Income Tax Rate, %
Escalation Rates, %
Cost of Capital
$14,300
$ 100
$ 500
$ 100
$ 0
$ 0
$ 800
$ 27
$ 500
100%
0%
10.00%
0
7
34.00%
5.0%
15.00%
OUTPUT
CAPITAL COST
Construction Year
Capital Expenditures
Equipment
Materials
Installation
Plant Engineering
Contractor/Engineering
Permitting Costs
Contingency
Startup Costs
Depreciable Capital
Working Capital
Subtotal
Interest on Debt
Total Capital Requirement
I
Equity Investment
Debt Principal
Interest on Debt
Total Financing
1
$14,300
$ 100
$ 500
$ 100
$ 0
$ 0
$ 800
$ 500
$16,300
$ 27
$16,327
$ 0
$16,327
$16,327
$ 0
$ 0
$ 1 6,327
i
'i
65
-------
TABLE F-2. NET OPERATING COSTS/SAVINGS FOR THE ATMOSPHERIC UNIT
DIFFERENCE IN OPERATING COSTS
(parentheses indicate negative values)
Marketable By-products
Total $/yr
Utility Cost
Electric
Water
Total $/yr
Raw Material Savings
Solvent
Water
Total
Decreased Waste Disposal
Spent Solvent
Residue
Labor
Transportation
Storage Drums $
Total Disposal $
$ 0
$ 0
$ 51
$ 18
$ 69
$ 6,670
$ 0
$ 6,670
$ 6,550
($3,220)
$ 40
$ 0
$ 480
S 3,850
Operating Labor Cost
Operator hrs
Wage Rate, $/hr
Total $/yr
Operating Supplies
Still bags
Maintenance Costs
Labor
Materials
Supervision Costs
(% of O&M Labor)
Overhead Costs
17
$ 8
$ 136
$ 120
$ 96
$ 86
10.0%
(% of O&M Labor + Supervision)
Plant Overhead
Home Office
Labor Burden
25.0%
20.0%
28.0%
66
-------
TABLE F-3. ANNUAL OPERATING SAVINGS FOR THE ATMOSPHERIC UNIT
REVENUE AND COST FACTORS
Operating Year Number
Escalation Factor
INCREASED REVENUES
Increased Production
Marketable By-products
Annual Revenue
1.000
1
1.050
$ 0
$ 0
$ 0
2
1.103
$ 0
$ 0
$ :0
3
1.158
$ 0
$ 0
$ 0
OPERATING SAVINGS (Numbers in parentheses indicate net expense)
Raw Materials
Disposal Costs
Maintenance Labor
Maintenance Supplies
Operating Labor
Operating Supplies
Utilities
Supervision
Labor Burden
Plant Overhead
Home Office Overhead
Total Operating Savings
$ 7,004
$4,043
($ 101)
($ 90)
($ 143)
($ 126)
($ 72)
($ 24)
($ 75)
($ 67)
($ 54)
$10,294
$ 7,354
$4,425
($ 108)
($ 95)
($ 150)
($ 132)
($ 76)
{$ 26)
($ 79)
($ 70)
{$ 56)
$10,80;8
$ 7,721
$4,457
($ 1 1 1.)
($ 100)
($ 157)
($ 139)
($ 80)
($ 27)
($ 83)
($ 74)
($ 59)
$11,349
67
-------
TABLE F-4. RETURN ON INVESTMENT FOR THE ATMOSPHERIC UNIT
RETURN ON INVESTMENT- -ATMOSPHERIC UNIT
Construction Year
Operating Year
Book Value
1
$16,300
Depreciation (by straight-line)
Depreciation {by double DB)
Depreciation
Cash Flows
1
$11,643
$ 2,329
$ 4,657
$ 4,657
2
$ 8,316
$ 2,329
$ 3,327
$ 3,327
3
$ 5,940
$ 2,329
$ 2,376
$ 2,376
Construction Year
Operating Year
Revenues
+ Operating Savings
Net Revenues
— Depreciation
Taxable Income
- Income Tax
Profit after Tax
+ Depreciation
After-Tax Cash Flow
Cash How for ROI
Net Present Value
Return on Investment
1
($16,327)
(16,327)
1
$ 0
$10,294
$10,294
$ 4,657
$ 5,637
$ 1,916
$ 3,720
$ 4,657
$ 8,377
$ 8,377
($9,042)
-48.69%
2
$ 0
$10,808
$10,808
$ 3,327
$ 7,482
$ 2,544
$ 4,938
$ 3,327
$ 8,265
$ 8,265
($2,793)
1.29%
3
$ 0
$11,349
$11,349
$ 2,376
$ 8,973
$ 3,051
$ 5,922
$ 2,376
$ 8,298
$ 8,298
($2,663)
24.64%
68
-------
APPENDIX G j
MAINTENANCE COSTS OF THE VACUUM UNIT
Based on information from the vendor, the following annual rciutirie maintenance costs
were estimated. The unit is expected to be operated for approximately 660 hr/yr at this site, based
on plant information and this testing, which showed 55 batches/yr and 12 hr/batch.
• Pump oil change once every 3 months, i.e., 4 times/yr and 6.5 hr/change.
• Ceramic filter cleaning once in 3 months (4 times/yr and 6.5 hr/change). Ceramic
filter changed once in 2 yr at $129.55/filter.
• Pump overhaul once in 2 yr (4 hr/overhaul). Spare parts {gaskets, 0-rings, etc.) are
$216/overhaul (maximum). '
j
• Cleaning inside of evaporator once a year (2 hr/yr). Spare Iparts are $307/cleaning
(maximum).
69
-------
APPENDIX H
ECONOMICS OF VACUUM UNIT
The costs utilized in the economic spreadsheets for the vacuum unit (Tables H-1 to
H-4) are described below:
• The cost of equipment of $25,900 includes the listed price of $23,500 plus 10% to
cover freight, taxes, insurance, and spare parts.
• Materials, installation, and plant engineering costs are estimates based on experi-
ence. Contingency costs are based on an assumption of 5% of fixed capital costs.
• Working capital is based on a 1-month supply of materials (drums). Startup costs
are estimated.
• Information on debt, depreciation, taxes, inflation, and cost of capital are general
assumptions.
• Differences in operating costs are based on costs listed in Table 10 of the main
report. Supervision and overhead costs are general assumptions.
The payback period can be determined from Table H-4 as the year in which the return on
investment (ROD exceeds 15% because 15% is the cost of capital (see Table H-1).
70
-------
TABLE H-1. CAPITAL COST OF THE VACUUM UNIT
INPUT
CAPITAL COST
Capital Cost
Equipment
Materials
Installation
Plant Engineering
Contractor/Engineering
Permitting Costs
Contingency
Working Capital
Start-up Costs
% Equity
% Debt
Interest Rate on Debt, %
Debt Repayment, years
Depreciation period
Income Tax Rate, %
Escalation Rates, %
Cost of Capital
$25,900
$100
$500
$100
$0
$0
$1,330
$10
$500
100%
0%
10.00%
0
7
34.00%
5.0%
15.00%
1
OUTPUT ! •
CAPITAL REQUIREMENT
Construction Year
]; .
Capital Expenditures
Equipment
Materials
Installation | .
Plant Engineering
Contractor/Engineering
Permitting Costs
Contingency
Start-up Costs
Depreciable Capital
Working Capital
Subtotal
Interest on Debt
Total Capital Requirement
Equity Investment
Debt Principal
Interest on Debt
Total Financing
1
$25,900
$100
$500
$100
$0
$0
$1,330
$500
$28,430
$10
$28,440
$0
$28,440
$28,440
$0
$0
$28,440
71
-------
TABLE H-2. NET OPERATING COSTS/SAVINGS FOR THE VACUUM UNIT
DIFFERENCE IN OPERATING COSTS
(parentheses indicate negative values)
Marketable By-products
Total $/yr.
Utility Costs
Electric
Water
Total $/yr.
Raw Material Savings
Solvent
Water
Total
Decreased Waste Disposal
Spent Solvent
Residue
Labor
Oil from recycling unit
Storage Drums $
Total Disposal $
$0
$0
$39
$0
$39
$9,830
$0
$9,830
$7,500
($340)
$104
($12)
$2,080
$9,332
Operating Labor Costs
Operator hrs
Wage rate, $/hr.
Total $/yr.
Operating Supplies
Maintenance Costs
Labor
Materials
Supervision Costs
(% of O&M Labor)
Overhead Costs
55
$8.00
$440
$0
$128
$494
10.0%
(% of O&M Labor + Supervision)
Plant Overhead
Home Office
Labor Burden
25.0%
20.0%
28.0%
72
-------
TABLE H-3. ANNUAL OPERATING SAVINGS FOR THE VACUUM UNIT
.... ,. , " _ .
REVENUE AND COST FACTORS
Operating Year Number
Escalation Factor
INCREASED REVENUES
Increased Production
Marketable By-products
Annual Revenue
1.000
1
1.050
$0
$0
$0
2
1.103
$0
$0
$0
3
1.158
$0
$0
$0
OPERATING SAVINGS (Numbers in parentheses indicate net expense)
Raw Materials
Disposal Costs
Maintenance Labor
Maintenance Supplies
Operating Labor
Operating Supplies
Utilities
Supervision
Labor Burden
Plant Overhead
Home Office Overhead
Total Operating Savings
$10,322
$9,799
($134)
($519)
($462)
$0
($41)
($60)
($184)
($164)
($131)
$18,426
$10,838
$10,289
($141)
($545)
($485)
$0
($43)
($63)
($1 93)
($172)
($138]
$19,347
$11,379
$10,803
($148)
($572)
($509)
$0
($45)
($66)
($203)
J$181)
J$145)
520,314
.4
1.216
$0
$0
$0
$11,948
$11,343
($156)
($600)
($535)
$0
($47)
($69)
($213)
($190)
($152)
$21,330
73
-------
TABLE H-4. RETURN ON INVESTMENT FOR THE VACUUM UNIT
RETURN ON INVESTMENT — VACUUM UNIT
Construction Year
Operating Year
Book Value
1
$28,430
Depreciation (by straight- line)
Depreciation (by double DB]
Depreciation
Cash Flows
Construction Year
Operating Year
Revenues
+ Operating Savings
Net Revenues
- Depreciation
Taxable Income
— Income Tax
Profit after Tax
+ Depreciation
After-Tax Cash Flow
Cash Flow for ROI
Net Present Value
Return on Investment
1
($28,440)
($28,440)
1
$20,307
$4,061
$8,123
$8,123
1
$0
$18,426
$18,426
$8,123
$10,303
$3,503
$6,800
$8,123
$14,923
$14,923
($15,464)
-47.53%
2
$14,505
$4,061
$5,802
$5,802
2
$0
$19,347
$19,347
$5,802
$13,545
$4,605
$8,940
$5,802
$14,742
$14,742
($4,317)
2.86%
3
$10,361
$4,061
$4,144
$4,144
3
$0
$20,314
$20,314
$4,144
$16,170
$5,498
$10,672
$4,144
$14,816
$14,816
$5,425
26.23%
4
$6,299
$4,061
$2,960
$4,061
4
$0
$21,330
$21,330
$4,061
$17,268
$5,871
$11,397
$4,061
$15,459
$15,459
$14,263
38.03%
74
-------
APPENDIX I
ENERGY REQUIREMENT OF LEVD
The energy requirement for a 1 -hour cycle is calculated in the
ratings of the heaters and pumps in the LEVD and their period of operation
Table 1-1, based on the
during each stage.
75
-------
I
D
m
o
z
g
i
CO
o
o
g
o
cc
§
LU
CQ
J2
>,
O
^_
£
•5
CD
IO
•^-
CO
CM
T""
0)
1
0
CO
1-
IO
cvi
CM
CM
CM
,_
in
'jc
£ £:
g"!
-io
CD
D) ^
I*
a
O £ CD "sT O CO **•
IB 5 t^ d CM d d
^j ^J _^
o •*
* *
O) IO O) "4"
3 d °"
* *
O) IO O ^^_^ &,
'r'ij •(!
Q. — s^.
">~ i_ o "tr CD
S D. w CD w
^, _ w = co
C a) •= Q)
i_ m »- -fr ir
0) ™ O-Q D-
"co 5. E c E
CD = O (0 O
X U_ O U_ O
1
CO
T—
CM
0
E
CD
•*-*
b c
CO O
co C7)''S
(D *- CD -S
O O t- ^3
° §-c?c3
* > Q O
II II II
T- CM CO
0) 0 CD
D) D) D)
CO CO CD
CO CO CO
Is
0
C Q.
0 3
'•<" *-i
1 S = c
lilt
0) ^ O O
>- i± CO CO
.b= CO 0) TJ
< O Q <
II II II II
^ U5 CO h-
0) CD CD CD
D) D) D) D)
CO CO CO CO
CO CO CO CO
76
-------
APPENDIX J
ECONOMICS OF LEVD
The costs utilized in the economic spreadsheets for the LEVD
are described below:
• The cost of equipment of $231,000 includes the listed price
unit (Tables J-1 to J-5)
of $210,000 plus 10%
to cover freight, taxes, insurance, and spare parts.
• Materials, installation, and plant engineering costs are estimates based on experi-
ence. Contingency costs are assumed at 5% of fixed capital costs.
• Working capital is based on a 1-month supply of materials (PCE). Startup costs are
estimated. |
• Information on debt, depreciation, taxes, inflation, and cosit of capital are general
assumptions. ;
• Differences in operating costs are based on costs listed iri Table 12 of the main
report. Supervision and overhead costs are general assumptions.
Because of the high capital cost of the LEVD, a loan of 60% of the cost was assumed.
Because the calculation for return on investment (Table J-4) does not take into account the costs
of a loan, a second spreadsheet Table J-5, which includes these costs, is {provided. This second
- l
spreadsheet includes both a return on equity (return on the 40% capital pbt up by the purchaser)
and a return on assets (return on the total capital). |
If no loan is taken (that is, if the purchaser puts up the entire capital himself), then the
payback period (year in which return on investment exceeds 15%) is 9 years {Table J-4). If 60%
of the capital is raised through a loan, then the payback period (year in which return on assets
exceeds 15%) is more than 10 years (Table J-5).
77
-------
TABLE J-1. CAPITAL COST OF THE LEVD
INPUT
CAPITAL COST
Capital Cost
Equipment
' Materials
Installation
Plant Engineering
Contractor/Engineering
Permitting Costs
Contingency
Working Capital
Start-up Costs
% Equity
% Debt
Interest Rate on Debt, %
Debt Repayment, years
Depreciation period
Income Tax Rate, %
Escalation Rates, %
Cost of Capital
$231,000
$500
$1,000
$1,000
$0
$0
$1 1 ,700
$1,060
$2,100
40%
60%
10.00%
7
7
34.00%
5.0%
15.00%
OUTPUT
CAPITAL REQUIREMENT
Construction Year
Capital Expenditures
Equipment
Materials
Installation
Plant Engineering
Contractor/Engineering
Permitting Costs
Contingency
Start-up Costs
Depreciable Capital
Working Capital
Subtotal
Interest on Debt
Total Capital Requirement
Equity Investment
Debt Principal
Interest on Debt
Total Financing
1
$231,000
$500
$1,000
$1,000
$0
$0
$1 1 ,700
$2,100
$247,300
$1,060
$248,360
$14,328
$262,688
$105,075
$143,285
$14,328
$262,688
78
-------
TABLE J-2. NET OPERATING COSTS/SAVINGS FOR THE LEVD
DIFFERENCE IN OPERATING COSTS
(parentheses indicate negative values) ;
Marketable By— products
Utility Costs
Electricity
Water
Electric heat
Total $/yr.
Raw Materials Savings
Solvent
Total
Decreased Waste Disposal
Spent Solvent
Residue
Labor
Transportation
Storage Drums $
Total Disposal $
$0
$3,749
($480)1
($1,0201
$2,249
$1,876
$1,876
$0
$0
$0
$0
$0
$0
Operating Labor Costs
Operator hrs
Wage rate, $/hr. !
Total $/yr.
Operating Supplies;
Chemicals
i
Maintenance Costs;
Labor
Materials
,t
Supervision Costs \
(% of O&M Labor)
Overhead Costs
(3,667)
$8.00
($29,336]
$0
$1,920
$2,012
10.0%
(% of O&M Labor + Supervision)
Plant Overhead
Home Office
Labor Burden
25.0%
20.0%
28.0%
79
-------
LU
H
CC
O
UL.
CO
O
I
a
1
ffi
o
_i
<
z
z
<
*?
-5
LU
m
^
o
L^i
I
1
CO
EC
R
f\
REVENUE AND COST FAC
00
f-
co
in
^
00
CM
T-
o5
.a
E
z
1
D)
1
a
C
•*
1
V™
§
CO
T~
o
M
"
CD
•vl
T~
s
§
8
o
T—
^
Escalation Factor
INCREASED REVENUES
e^
»
»
»
o
«•
o
&•
o
«a
o
01
n
JC
S
0
.Q
E
i
OPERATING SAVINGS (Nt
N
CM
tfr
§
CM
<#
-3-
10
CM
ie-
•*
en
CO
CM
*3-
§
CM
<&
CM
N.
CM
<£•
00
cS
CM
<«•
O
&
1&
Raw Materials
&
&
&
&
&
&
&
o
te-
•ft
o
O
"cc
a
c
V)
b
r>
CO
CM
^
CM
D
CM
^
X)
n
CM
^
n
TJ-
CM
^
•*
CO
CO
CM
«£
C^
CM
CM
S£
N.
CM
$&
(O
T—
O
CM
£i
Maintenance Labor
T5
3>
CM
&
5
00
CM
%t
CD
O5
CD
CM
^
00
CD
ID
CM
^
CD
•^J-
•"3-
CM
«>
0)
CM
CO
CM
£t
00
CM
CM
^.
CO
T—
CM
^.
Maintenance Supplies
CO
oS
CO
•*
*&
3>
•M
5-
«3-
CO
•o
0)
CO
<«•
5
7)
*»
00
LO
CD
to
CO
«/3-
O
CO
O>
CO
CO
*»
CO
•^r
CO
CM-
CO
i»
CO
o
CO
o
CO
«•
i_
o
n
cc
C33
c
"I
0
Q.
o
»
&
fir
A-
»
•&
o
*&
o
f&
to
CD
"o.
c
v:
DJ
1
0
Q.
o
s
CO
CO
&
LO
o
CO
^
•*•
b
CO
%t
ID
X)
CM
*5.
•«*•
TO
CM
£i
CO
o
CD
IN
f£
§
"*•
CM
^
f—
CD
CO
CM
^
-I
s
§
•t
te^
00
o
00
CO
ta-
•*
D
CO
«•
O)
9
CO
1&
CM
CO
CO
CO
•66-
"f
^
CO
00
CM
CM
CO
00
&
CO
CO
5»
^
o
»
•5J-
$
o
^_
<«•
JO
s.
o
*&
o
CO
o>
t»
CO
§
CO
CO
«e-
CO
s
r>.
«3-
§
Is.
<«•
CM
§
CD
-------
CC
o
at
LU
CC
i
cc
t
-J
LU
p
LLJ
1
1
1
III
••»
CO
m-
o
r
ID
n
*-
CO
U)
5
Construct
CO
^
CO
in
*
CO
CM
*~
1
Operating
o
w
o
if.
N.
3
O)
in
0)
h-
m
<&
§
o
o
73
ft
«*
CO
M-
CD
D
T —
ttf
D
CO
!*••
^Xl
&•
CD
B
«:
Q
O
W
&
O)
CV
CO
in
(.0
*ni
0)
CM
CO
in
CO
CV
CO
in
CO
^
O)
CM
CO
in
CO
O)
CM
CO
•n
ib
in
CO
CO
0)
c
^
O)
ta
w
5
io
5
5.
Q
o
<»
if
CM
£
in
w-
CO
m
CO
in
o
[2
in
CM
<#•
§
o
CD
CO
0)
CO
•<3-
o
n
n
CO
D
•CO
n
CD
-Q
_J
o
TJ
^i
5
S
L>
5.
Q
o
w
o
CD
CD
T~
\jf
O)
CM
CO
m
cc
O)
CM
CO
m
CO
<»
0
in
0
co~
CO
O)
'£>
n
n
CO
D
^-
5
S
3
5.
Q
U)
O
u_
i
•1-
(Tl
CD
5
Construct!
CO
N-
CO
in
*
CO
CM
•«~
5
D
Operating
o
w
o
o
c
o
•w-
p
p
p
Revenues
0)
u
m
CO
N.
CD
in
o
o
fs^
w-
r^*
CD
CO
^"
5
in
CO
cc
in
0
CD
tw-
in
CO
CO
n
X)
n
M
CM
in
«•
O)
fc
CO
O)
=
D
Q.
O)
in
in
CO
r*-
co
o
o
f*^
t»
cv
N.
CO
CD
^f
CO
in
CO
cc
in
o
CD
tw-
in
CO
CO
LO
0)
00
n
M
CM"
10
ft"
U)
D
5
5
o
«>
o
N.
CD
O)
T^
^Ji
O)
C\
CO
in
cc
O)
CVJ
CO
in
CO
w
o
n
CO
CO
*
O)
D
n
n
CO
D
£•
5
0
i)
i
D
1
O)
in
in
CO
CO
o
o
f^S,
to-
co
c3
fs.
5^
^f
T—
CM
CO
CV
O)
CO
m
CM
*«•
CO
00
in
CM
w
§
S
D
CO
CO
/?"
come
»
0
0
o
in
CV
O)
00
CO
CV
CD
CC
CO
^f"
cc
a
i
cc
O)
CO
CO
p>-
o
£
O)
CM
D
"^•* *
B
- Income
O)
•<*
in
CO
N.
CM
CO
^^
g f
p1^
CO
,_
cc
•ff
CM
CV
CO
CO
s
CD
CO
W
§
*
*w
CO
a>
:M
!o
CM"
f^.
1
0
5
L
O
O
CD
O)
£f&
O)
CM
CO
in
CO
O)
CM
CO
in"
CO
w
g
CO
CO
**!
O)
D
n
n
CO
D
A
i
5
D
D
0)
s
00
N.
CM
CO
^-
**
in
CD
o
in
u:
O)
CO
1C
CO
in
O)
in
«>•
CO
O)
CM
o
10
tw-
CO
CO
CO
CO
in
CD
CM
in
CO"
in
o
LJL
1
O
1
1
O)
in
CO
N-
CM
CO
**
S
CD
o
in
ff
IT
O)
CO
LC
CO
in
O)
in
£>
0
w
CO
CO
CO
CO
in
CD
CM
in
co~
in
CO
on
•<*•
CM
n
cc
5
D
i_
1
O
CD
or
cv
CV
CO
rv
c\T
§
in
ff
— •
o
in
cc
'•^•^
DC1
O)
*^ -"
C^
cti
s
§
n
&
S
3)
*a
§
CO
CO
^r
^
CD
0
n
L
D
s<
c\i
co'
T—
s^
O
'""
K.
h*»
i
cvi
S
in
1
O)
1
00
o"
1
n
co"
1^*
1
*•*
S
D
5
D
81
-------
fl
1
1
CO
CO
rn
*d
kM
,__
rr*
I
Q
^^
^
«2.
o
2r
cc
t
T-
(
"c
1
c
C°
X)
•s.
CD
in
*
CO
CM
*-
C
c
D
C
1
CO
a
O
3>
o
ee
D
of
e-
33
I
t\i
~
2
£
£
CO
"C
(O
ID
,—
te-
2
IT)
r**
r\
c
^
o
o
CO
D
j}
O
J
ro
3
n
fl
cn
CM
CO
c
14
cn
CM
ro
in
S
i
r
r
•69-
7
Y
LC
•69-
C
.t=
•*
D
L_
CO
1
c
<
1
o
CD
a
i
Q
D
j^.
O
CM
CD
in
3
3
1
iri"
D
\
r
CD
1
r
r
-
^
CD
n
^
y;
_i i
c
Q)
E
(
Q
C
C
Tfi
C
CO
C
Q.
U
O
Y-
c
(
c
g
c
i
c
o
O
X)
CO
in
"*
CO
CM
••-
i.
(
C
D
C
1
CU
a
O
o
-«•
o
FA-
D
O
D
f.
f.
f.
CO
(
(1
i
CC
33
O
3
•o
A
CD
n
o
D
09-
M
3
3-
CO
in
o
9
-
£
in
£
n
fA-
7
in
^
•y
in
•te
c
c
c
CO
a
j
i
i
a
O
33
n
3
T>
e-
CO
n
o
D
69-
M
CD
3
CO
in
9-
in
(3
r
£
n
j
•«*•
in
^
Jr
w-
c
c
CD
CC
1
2
D
•ft
FA-
3
cn
cn
CM
co
b
cn
CM
CO
in
rj
r
CD
tr\
7.
C.
n
CO
5
te
•eciation
0
Q
D
y*
"VI
CM
FA-
CO
b9-
•3
O
CD
7)
CO
c
1—
f\
in
CO
D
in
T™
W-
C
Q
c
o
*
CO
(
n
n
T)
=9-
in
CO
o
69-
O
3
CM
te-
3
M
CM
D
CD
CU
CM
r
o
f
CO
7)
5
CO
w-
0
E
o
(
X
i
t-
3
_p
O
n
M
=e-
o
CO
CM
ys»
3
te
CD
»
\
s
r
5
CO
3
CO
•te-
X
I
*~
•fey
i
t-
i
i
i
•
jO
DO
)•
7>
p
^
CM
M
te-
rj
•sf
o
D
0
ID
CO
D
D
3
r
CO
Y
m
£
f^
i
h-
c
j^
03
o
i
_3
&
FA
D
cn
5>
cn
CM
CO
LO
CO
te-
en
CM
CO
in
CO
•ee
r
CO
f
cn
o
r
^3>
D
in
N
w-
CD
LO
e-
n
CM
M
U3
CM
\
CD
LO
CM
(M
f
CD
LO
Cv
CD
n
S^
•ty-
4*(
<
E
(
Q
C
CC
XI
1
£J
^
)
33
^
O
CM
M
b9
LD
3
N
CD
33
N
7)
M
C
rv
f-
in
33
CXJ
Z
r
•ty-
^J
c
LL.
C
C
C
1—
%
<>
J
33
^>
CO
CM
CM
CM
by-
(D
3
M
fee
CD
3)
M
33
PJ
^
IT
•v
f
in
3
CO
£
IT
•ty-
u
o
in
t—
{£.
n
O
CC
u.
H—
j
u_
SL
U
1
O
N
3)
**•
c^
J
in
a
13
M
i
cn
f.
LO
1C
CM
in
f
i\
CO
CD
CU
X
a
•w-
If
o
in
c
t—
i2
<
T
"c
c
CO
a.
•*»J
1
^L
•
^
CO
cvi
3
n
cn
D
"*
vP
in
1
.c
(O
O3
1
In
o
1
:c
n
1
u
c
o
1
H
1
cc
)
X3
b9-
O
>4
CM
by-
ID
3
N
CD
33
M
CO
33
N
LO
!v
f.
U)
7
CU
I
u:
-------
APPENDIX K
LEVD CYCLE TIME VARIATION FOR VARIOUS
METALS
Table K-1 lists the thermal conductivity (k), density (d), and J
as given in Chemical Engineers' Handbook (Perry and Chilton, 1973) for
and copper. The thermal diffusivity (a) is calculated from these values by
ipecific heat (Cp) values
steel, aluminum, brass,
tr e relationship
Based on experimental total cycle times measured for 16!5 and 915 Ib of steel
(Table 11 in this report), the total cycle time for 1,100 Ib of steel (maximum load) is calculated by
linear extrapolation. From the total cycle times for steel, 40 min (the constant component) is
subtracted to give the variable component of the vapor-fill stage time. The relationship
ametal • tmetal = asteel • tsteel is then applied to the variable component of vapor-fill times for
steel to obtain the corresponding times for the other metals.
83
-------
TABLE K-1. VARIATION OF LEVD CYCLE TIMES FOR VARIOUS METALS
Steel
Aluminum
Brass
Copper
k
(Btu/ft/hr/FL
26
118
58
220
Density
(Ib/ft3)
490.8
169.3
520.1
556.4
Specific Heat
(Btu/lb/F)
0.107
0.23
0.09
0.092
Thermal
Diffusivity
(ft2/hr)
0.495
3.030
1.239
4.298
Steel
Aluminum
Brass
Copper
Variable
Component
of
Vapor
fill Total
(min) (min)
11 51
2 42
4 44
1 41
Mass
(Ibs)
165
57
175
187
Variable
Component
of
Vapor
fill Total
(min) (min)
29 69
5 45
12 52
3 43
Mass
(Ibs)
915
316
970
1037
Variable
Component
of
Vapor
fill Total
(min) (min)
34 74
g
14 54
4 44
Mass
(Ibs)
1100
379
1166
1247
84
U.S. GOVERNMENT PRINTING OFFICE: 1994 - 550-001 / 80363
-------
-------
EPA/600/R
co
ro
o
Official Bus
Penalty for
$300
135-
~
CD (Q
W CD
CD 3
W O
3^
3_
o"
—\
21
o'
E? o ' 5T Q)
?g.c jjTg.!
So ° ^-^p
O O CD C/l
S 8 ? S
-- < => >2
§S °s-
°-i" s-§
fn W O> ^
W =r
=•»
ix ?E
-D
O
-0 C/3
m H.
33 O c
— _ m i—
>
o
------- |