MEMORANDUM

DATE:		April 13, 2007

SUBJECT:	National Impacts for Final Rule  

FROM: 	Chris Sarsony

		engineering-environmental Management (e2M), Inc.

THROUGH:	John Crenshaw

		Eastern Research Group, Inc.

TO:		Lynn Dail

		U.S. Environmental Protection Agency

1.0	INTRODUCTION

On December 2, 1994, EPA promulgated technology-based emission standards
to control HAP emissions of halogenated solvents from halogenated
solvent cleaning.  Pursuant to the Clean Air Act (CAA) section 112(f),
EPA evaluated the remaining risk to public health and the environment
following implementation of the technology-based rule and determined
that more stringent standards are necessary in order to provide an ample
margin of safety to protect public health.  

On August 17, 2006 EPA published proposed standards that provided
further reductions of methylene chloride (MC), perchloroethylene (PCE),
and trichloroethylene (TCE) beyond the 1994 national emission standards
for hazardous air pollutants (NESHAP), through application of a
facility-wide total MC, PCE, and TCE emission standard.  The EPA
received comments on the proposed rule from industry, states, solvent
manufacturers, industry associations and district air associations. 
Industry’s comments were primarily submitted by four specific sectors:
narrow tubing manufacturing facilities, facilities that manufacture
specialized products requiring continuous web cleaning, aerospace
manufacturing and maintenance facilities, and military depot maintenance
facilities.  Comments and data submitted by the four industry sectors
focused on the unique nature and size of the halogenated solvent
cleaning machines used in the cleaning operations.  

A Notice of Data Availability (NODA) was issued on December 14, 2006 (71
FR 75184) to collect additional information from these four industry
sectors.  As a result of the NODA, EPA received significant comments
from responders associated with the above-noted industries, industry
associations, and commenters that were not associated with the
above-noted industries.  They provided additional data and information
on control costs and the technical feasibility of additional controls
that were directly relevant to the promulgation of the proposed
facility-wide emission limits.  

The purpose of this memorandum is to summarize the nationwide cost and
emission impacts associated with the control options evaluated during
the development of the final rule.  In Section 2.0 the changes to the
impacts analyses that resulted from the public comments received are
summarized.  Section 3.0 contains a description of the 2002 degreasing
database.  Section 4.0 contains a discussion of the levels and format of
the control options.  Section 5.0 summarizes the development of costs
and percent emission reduction for individual control techniques. 
Section 6.0 explains how the affected facilities were identified and
describes the methodology for applying the individual control costs to
the solvent cleaning facility database for each of the control options. 
 Section 7.0 summarizes the results of the cost analysis for all of the
control options.



CHANGES TO IMPACTS METHODOLOGY RESULTING FROM PUBLIC COMMENTS

Thirty-eight comment letters were received in response to the August 17,
2006 proposal and an additional twelve comment letters were received in
response to the December 14, 2006 NODA.  Included in these comment
letters were data and information on control costs and the technical
feasibility of additional controls (See Comment & Response Document and
the NODA Comment & Response Document for a full summary of the comments
received and EPA’s responses to them.).  The EPA evaluated the
information provided in the comment letters and contacted some
commenters for additional discussions and clarifications of their
comments where necessary.  Based on the analysis of the comments
provided by industry EPA determined that the changes summarized in Table
1 would be incorporated into the impacts analysis for the final rule.   

In the proposal analysis if a unit were switching from PCE to TCE, TCE
to MC, or TCE to MC the costing spreadsheet did not take into account
the difference in the cost of the initial solvent versus the new
solvent.  Since PCE is the most expensive solvent, TCE is the second
most expensive, and MC is the least expensive, the approach used
previously underestimated the solvent savings.  This was corrected by
restructuring the cost spreadsheet to show columns for the initial
solvent and the new solvent.  The solvent savings are then calculated as
the difference between the amount of initial solvent used x the cost per
gallon for that solvent minus the amount of the new solvent used x the
cost per gallon for the new solvent.  The change more accurately
reflects the changes in costs when switching from one solvent to
another.  

The other aspect of the solvent switching calculation that was adjusted
was the method that was used to calculate the emission reductions.  For
example, at proposal, it was assumed for PCE to TCE that emissions would
be reduced by 77%.  Therefore, the spreadsheet showed a reduction of 77%
in PCE emissions.  However, a switch from PCE to TCE would result in no
PCE being used for that unit because it would be replaced with TCE.  The
magnitude of TCE emissions would be 31% less than the PCE emissions for
that unit [See memorandum in the docket titled “Evaluation of the
Feasibility, Costs, and Impacts of Switching from a Halogenated Solvent
with a High Cancer Unit Risk Value to a Halogenated Solvent with a Lower
Cancer Unit Risk Value”]   The additional reduction in risk is not the
result of less emissions, but is a result of the lower MIR risk of TCE
relative to PCE.  Table 2 below summarizes emissions reductions and risk
reduction values for each of the solvent switching options.  

Table 1 - Changes to Cost and Emissions Impacts Methodology Since
Proposal for Main Population of Degreasers

Change	Effect	Notes

The percent emission reduction associated with vacuum to vacuum (VTV)
cleaning machines was reduced from 97% to 95%.	This decreased the amount
of reduction achieved.	This change was made as a result of public
comments. 

The VTV cleaning machine O&M costs were increased from $0 to $18,832.  
This increased costs.	The O&M costs for VTV cleaning machines were
estimated to be half of annualized capital costs.  Prior to proposal a
manufacturer of VTV cleaning machines indicated that there typically
would be no increase in electricity and labor costs.  However, costs
obtained by tube and aerospace manufacturers from the same VTV
manufacturer indicated that there would be an increase in O&M costs.

Carbon adsorption (CAD) was added at 30% control.	Increased costs
because this control is 10 times more expensive than other controls
options at 30%.	This control was added because information received from
tube manufacturers, aerospace manufacturers, and users of web cleaning
machines indicated that CAD was one of the few a viable control option
for their cleaning machines.

Cost per gallon of PCE and TCE were updated.  PCE was changed from
$11.10/gallon to $14.23/gallon.  TCE was changed from $12.83/gallon to
$12.74/gallon.	Overall effect was a decrease in costs because the
solvent savings credit increased.	Updated costs used were based on
commenter submitted costs.

The percent of total affected units that were assumed to choose solvent
switching to comply with the rule was reduced from 30% of the affected
units to 15% of the affected units.	Reducing the number of units
choosing solvent switching increased the population costs because the
solvent switching options are lower in cost than other options.	Change
made in response to comments.

The methodology that was used to calculate the emission reduction
impacts and solvent savings for solvent switching was corrected.	This
reduced the emission reductions and had a mixed effect on the solvent
savings.	The new methodology more accurately reflects the true effects
of solvent switching. 

Facility unit and emissions data were updated from 1999 NEI to 2002 NEI
Decreased emission reductions.	Since emissions have gone down there are
less emissions above a given level.  

Tube manufacturers, Aerospace manufacturers, facilities using Web
cleaning machines and the Anniston Army Depot were separated from the
rest of the cleaning machine population.	This decreased the emission
reductions. 	This change was made in response to public comments.



Table 2 – Change in Solvent Volume, MIR Cancer Risk, and Overall Risk
for the Solvent Switching Options

Solvent Switching Option	Change in Solvent Volume	Change in MIR Cancer
Risk	Overall Change in Risk

PCE to TCE	31% decrease	66% decrease	77% decrease

TCE to MC	29% increase 	76.5% decrease	70% decrease

PCE to MC	11% decrease	92% decrease	93% decrease



3.0   DEGREASING DATABASE

As indicated in Table 1 several significant changes were made since
proposal to the database upon which costs and emission reduction
estimates are based.  At proposal, the impacts estimate was conducted on
a database developed largely from EPA’s 1999 National Emissions
Inventory (NEI) database.  After the proposal was developed a final
version of the 2002 NEI database was released.  In response to public
comment and to ensure that the latest available information was
utilized, a new 2002 degreasing database was developed from the final
version of the 2002 NEI database (Final Version 2, Dated 10-23-06). 
This 2002 degreasing database was used as the basis for the impacts
analyses for the final rule.  A complete discussion of the development
of the 2002 degreasing database is discussed in the memorandum titled
“Development of the 2002 degreasing database from the 2002 NEI
Database” which can be found in the rule-making docket.  

	Facility and unit specific data from the 2002 degreasing database
served as the basis for estimating cost impacts.  Specifically, unit
specific data on PCE, TCE, and MC emissions and cleaning machine design
were used to determine which facilities were above the potential control
options and which specific controls could be applied to achieve each
control option limit.  A detailed description of how costs were applied
to each facility is contained in Section 6.0 of this memorandum.  

The 2002 NEI database contains data for 1,080 facilities.  The 1,080
facilities represent approximately 57 percent of the 1,900 total
facilities estimated to be in the source category.  Therefore, data from
the 2002 degreasing database have been scaled-up proportionally to
reflect results for the 1,900 facilities in the source category by
multiplying the number of units, facilities, and emissions by a factor
of 1.7 (1,900 facilities/1,080 facilities = 1.7).  Tables 3 and 4 show
population and emissions data for the 2002 degreasing database and the
scaled-up 2002 national population.  

Table 3 – 2002 Degreasing Database and the National Population

Category	2002 Degreasing Database	National Population Based on 2002
Degreasing Database

Number of Records	1,946	-

Number of Units	1,725	3,035

Number of Facilities	1,080	1,900



Table 4 – Halogenated Solvent Emissions from the 2002 Degreasing
Database and the National Population

Halogenated HAP Emissions	2002 Degreasing Database (tons/yr)	National
Population Based on 2002 Degreasing Database (tons/yr)

PCE (CAS# 127184)	889	1,564

Carbon Tetrachloride. (CAS# 56235)	0	0

Chloroform (CAS# 67663)	5	9

TCA (CAS# 71556)	2,300	4,046

MC (CAS# 75092)	567	997

TCE (CAS# 79016)	2,760	4,855

TOTAL	6,521	11,471



In addition, the results of the impacts analysis have been scaled-up
similarly to reflect results for the 1,900 facilities in the source
category by multiplying the costs and emission reductions by a factor of
1.7.

The second significant change regarding how the degreasing database was
utilized for the impacts analysis was that tube manufacturers, aerospace
manufacturers, facilities using web cleaning machines, and the Anniston
Army Depot were separated from the rest of the cleaning machine
population.   The cost and emission impacts were evaluated separately
for these facilities.  

Aerospace facilities were identified primarily on the basis of the
following SIC codes (3720, 3721, 3724, 3728, 3760, 3761, 3764, 3769,
4512, and 4581) and NAICS codes (336411, 336412, 336413, 336414, 336419,
481111, and 481112), which are listed in the Aerospace NESHAP. 
Facilities with web cleaning machines were identified from comments
provided on the proposal of the original NESHAP, comments on the
proposed residual risk standards, and other sources.  Facilities
manufacturing narrow tubing were identified from comments on the
proposed residual risk standards, comments on the NODA, and by searching
the facility name field of the 2002 degreasing database on the keyword
“tube”.  Anniston Army Depot was identified based on comments
submitted on the proposed residual risk standards.



CONTROL OPTIONS

Table 5 shows the control options that were evaluated for the final
rule.  All of the options evaluated are beyond the 1994 MACT levels. 
The control options for the main population, aerospace facilities, and
Anniston Army Depot are in the form of facility-wide solvent cleaning
emission limits that apply to all PCE, TCE, and MC emissions from
solvent cleaning operations at a facility.  Some facilities use only use
one of these solvents, but others may use two or all three.  The control
option evaluated for facilities with web cleaning machines is an
increase from 70% overall control efficiency to 80% overall control
efficiency for solvent cleaning operations.  The control option
evaluated for tube manufacturers was a 10% reduction in emissions from
solvent cleaning operations.

Table 5 - Emissions Control Options Evaluated for Final Rule

Solvents Emitted	Control Options – kg (lbs)

	Main Population	Aerospace	Anniston	Web	Tube

PCE only	1,992 (4,392)	3,187 (7,027)	4,781 (10,540)	7,968 (17,567)	7,968
(17,567)	19,920 (43,916)	7,968 (17,567)	Increase from 70% Overall
Control Efficiency to 80% Overall Control Efficiency	10% Emission
Reduction

TCE only	5,869 (12,938)	9,390 (20,700)	14,085 (31,051)	23,474 (51,751)
23,474 (51,751)	58,685 (129,378)	23,474 (51,751)



MC only	25,000 (55,115)	40,000 (88,183)	60,000 (132,276)	100,000
(220,460)	100,000 (220,460)	250,000 (551,150)	100,000 (220,460)



Multiple Solvents - MC Equivalent 	25,000 (55,115)	40,000

(88,183)	60,000 (132,276)	100,000 (220,460)	100,000 (220,460)	250,000
(551,150)	100,000 (220,460)





To allow the simultaneous assessment of risk from PCE, TCE, and MC
emissions, which have different cancer unit risk estimates (UREs) as
shown in Column two of Table 6, the control options were developed based
on the relative cancer potency of the HAPs emitted.  Since MC has the
lowest cancer unit risk it was used as the base risk and the risk values
for PCE and TCE were scaled relative to the MC unit risk estimate. 
Therefore, as shown in column three of table 1, the MC unit risk factor
is assigned a scaling factor of 1, the TCE unit risk factor is assigned
a factor of 4.26, and the PCE unit risk is assigned a factor of 12.55. 
This is simply showing that the cancer unit risk for TCE emissions is
4.26 times higher than the cancer unit risk for MC emissions, and that
the cancer unit risk for PCE emissions is 12.55 times higher than the
cancer unit risk for MC emissions.   

Table 6 - Cancer Unit Risk Estimates and MC Equivalent Factors

for TCE, PCE and MC

Halogenated HAP Solvent	Cancer Unit Risk Estimate (ug/m3)-1	MC
Equivalent Factor (Cancer Unit Risk Estimate of Pollutant/Cancer Unit
Risk Estimate of MC)

TCE	2.0E-06	4.26

PCE	5.9E-06	12.55

MC	4.7E-07	1



	The limits in the PCE only, TCE only, and MC only rows of Table 5 apply
just to those facilities that emit only one of the solvents.  Facilities
that emit multiple solvents must determine the weighed MC equivalent
emissions using equation 1 and meet the limits in the shaded row.  For
simplicity, the control options were expressed only in MC equivalents in
the cost summary tables.

Equation 1:    

MC Equivalent = (PCE emissions x 12.55) + (TCE emissions x 4.26) + (MC
emissions)

Emissions

	For example, for one control option, each facility must limit the total
combined emissions of PCE, TCE, and MC in kg to 60,000 kg MC equivalent.
 So for a facility that emits 4,000 kg of PCE, 2,000 kg of TCE, and
10,000 kg of MC the MC equivalent emission would be determined as
follows:

MC equivalent emissions in kg  

= (4,000 kg emissions of PCE x 12.55) + (2,000 kg of TCE x 4.26) +
(10,000 kg of MC)

= 50,200 kg + 8,520 kg + 10,000 kg

= 68,720 kg 

Therefore, this facility is 8,720 kg MC equivalent over the limit.  To
comply with the limit the facility could change practices or apply
controls to do the following:

Reduce PCE emissions by 695 kg,

Reduce TCE emissions by 2,047 kg,

Reduce MC emissions by 8,720 kg,

Or any combination of reducing PCE, TCE, and MC emissions where:

(PCE emissions reduction in kg x 12.55) + (TCE emissions reduction in kg
x 4.26) + (MC emissions reduction in kg) = 8,720 kg or more

Establishing the control options on a facility-wide basis allows each
facility the flexibility to comply in the most cost effective manner. 
This is because each facility can choose which units to control, which
controls to apply, and which solvents to control so long as the limit is
met.  

5.0  DETERMINATION OF COSTS FOR INDIVIDUAL CONTROLS

	At proposal, a suite of controls was developed that achieve emission
reductions beyond the level of the MACT and that reduce the level of
cancer risk associated with the emissions.  As a result of public
comments received on the proposal and on the NODA some costs were
modified and some additional controls were added.  Table 7 shows the
suite of controls and associated costs that were used for evaluating the
cost impacts of the final rule on the main population of cleaning
machines.  Table 8 shows the controls that were applied to the web,
tube, aerospace, and military depot cleaning machine populations.  

Table 7 - Controls Beyond MACT Applied to Main Population 

Control Type	Description	% Emission Reduction (a) 	Total Capital Costs
Annualized Capital Costs	O&M Costs	Total Annual Emission Control Costs
(b)

Control Equipment Retrofits	1.5 Freeboard Ratio (1.0FBR), Working Mode
Cover (WC), Freeboard Refrigeration Device (FRD)	50	$25,645	$2,821
$2,015	$4,836

	1.5 Freeboard Ratio (1.5FBR)	30	$20,380	$2,242	$0	$2,242

	Carbon Adsorption System (CAD)	30	$162,687	$17,896	$8,948	$26,844

Solvent Switching	PCE to MC	11	$15,677	$1,725	$928	$2,653

	PCE to TCE	31	$0	$0	($2,022)c	($2,022)

	TCE to MC	(29)	$15,677	$1,725	$2,950	$4,675

Machine Replacement	Vacuum to Vacuum Cleaning Machine	95	$399,000
$37,663	$18,832	$56,495

a -  For the solvent switching options the percent emission reduction
reflects the amount of new solvent emitted relative to the amount of old
solvent emitted.  For the TCE to MC option it is estimated that
emissions of MC would be 29% more than the original TCE emissions.  The
primary benefit for the solvent switching options results from the lower
MIR cancer risk for the new solvents.  If the MIR values are taken into
account along with the change in emission reduction, the resulting
decrease in cancer risk are as follows:

	PCE to MC – 93%

	PCE to TCE – 77%

	TCE to MC – 70%

b – Does not include cost savings due to reduced solvent purchases. 
The solvent savings were calculated for each specific unit based on the
volume of solvent emissions reduced and the cost of the specific solvent
in $/gal. 

c – Values in ( ) indicate a cost savings.

Table 8 - Controls Beyond MACT Applied to the Web, Tube, Aerospace, and
Military Depot Cleaning Machine Populations

Industry Segment	Description	% Emission Reduction	Total Capital Costs
Annualized Capital Costs	O&M Costs	Total Annual Emission Control Costs
(a)

Aerospace	Vacuum Machine - Small - Baskets	95	$150,000	$14,159	$7,080c
$21,239

	Vacuum Machine - Small - Hanging	95	$242,000	$22,867	$11,434c	$34,301

	Vacuum Machine - Large - Baskets	95	$584,000	$55,125	$27,563c	$82,688

	Vacuum Machine - Large - Hanging	95	$788,800	$74,457	$37,229c	$111,686

	Enclosure, parts transporter, CAD, and distillation	50	$237,500	$26,076
$13,038c	$39,114

	Dual-Split Working-Mode Cover and Dual-Bank Freeboard Chillers	25
$44,000	$4,831	$2,416c	$7,247

	Enclosure	15	$108,000	$11,858	$5,929c	$17,787

Anniston Army Depot	RETRO - 1.5FBR, WC, FRD (b) 	50	$102,581	$11,284
$8,059	$19,343

Tube	Side Chamber	20	$200,000	$22,000	$11,000c	$5,120

Web	Automated Gates AND Lengthening the cleaning process area, providing
an additional enclosure around all distillation units, and modifying the
capacity and process structure of the CADs	40	$850,000	$93,500	$46,750c
$123,003

a – Does not include cost savings due to reduced solvent purchases. 
The solvent savings were calculated for each specific unit based on the
volume of solvent emissions reduced and the cost of the specific solvent
in $/gal. 

b – To account for the large size of the cleaning machine the costs
are 4 times those used for the main population).

c – O&M costs were assumed to be equal to half the annualized capital
costs.

	The costs for the retrofit controls in Table 7 were based on vendor
estimates obtained in 2005 (Table A-1 and Table A-2).  The capital costs
were based on equipment for a solvent cleaning machine with a
solvent-air interface area of 2.5 m2, which is the average size of the
solvent cleaning machines in the database for which size data are
available.  The annualized capital costs were based on a 15 year
equipment lifetime and a 7% interest rate.   A 50% emission reduction is
expected to result from the addition of the 1.0FBR, WC, and FRD control
combination.  A 30% emission reduction is expected to result from the
addition of a 1.5FBR.  These percent emission reductions were calculated
using percent reduction values and procedures that were developed for
the NESHAP.

	The development of the costs for the solvent switching options shown in
Table 7 included considerations of changes in the cost of the solvent,
changes in solvent consumption rates, changes in energy requirements,
costs for equipment modifications, and changes in productivity.  Capital
costs were annualized assuming a 15-year equipment lifetime and a 7%
interest rate.  The solvent switching scenarios, their costs, and
impacts are fully discussed in a separate memorandum titled
“Evaluation of the Feasibility, Costs, and Impacts of Switching from a
Halogenated Solvent with a High Cancer Unit Risk Value to a Halogenated
Solvent with a Lower Cancer Unit Risk Value.”    

	Costs for the vacuum-to-vacuum cleaning machine in Table 7 are based on
vendor estimates obtained in 2005.  The vacuum-to-vacuum cleaning
machine capital costs were based on the replacement of a solvent
cleaning machine with a solvent-air interface area of 2.5 m2, which is
the average size of the solvent cleaning machines in the database for
which size data are available.  Since vacuum-to-vacuum cleaning machines
do not have a solvent-air interface, it was necessary to correlate the
solvent-air interface area of the old machine to the cleaning capacity
of the new vacuum-to-vacuum cleaning machine.  This was done using the
following relationship, which was originally developed for the NESHAP
and is presented in Docket A-92-39, item IV-B-2:

Solvent-Air Interface in m2  = 2.2 (Cleaning Volume in m3)0.6

Capital costs were annualized based on a 20 year equipment lifetime and
a 7% interest rate.  The 20-year equipment lifetime was determined based
on information from equipment manufacturers.  At proposal a 97%
reduction in emissions was assumed to result from switching from an
existing solvent cleaning machine to a vacuum-to-vacuum cleaning
machine.  However, based on comments received after proposal the assumed
emission reduction was changed to 95%.  

	The costs in Table 8 were derived from comments provided by affected
facilities in these industry sectors.  Additional details regarding the
cost information received from industry can be found in the worksheets
titled “Cost info from comments” and “Options info from
comments” that are located in docket in the cost spreadsheet for the
final rule titled “4-02-07 Costs”.

6.0 DETERMINATION OF AFFECTED FACILITIES AND COSTS

	For the main population, aerospace, and the Anniston Army Depot, the
affected facilities (i.e., the facilities that must reduce emissions)
were identified from the 2002 degreasing database based on whether the
facility total MC equivalent emissions in kg exceeded the control option
limit being evaluated.  The MC equivalent emissions were calculated for
each facility using equation 1 (shown in Section 3.0).  

	Once the population of affected sources was identified the required
controls and the resulting emissions reductions were calculated as
follows:

To determine the percent control required the following equation was
used:

(Facility MC EQ – Limit in MC EQ)/ (Facility MC EQ) = % Control

Required

A control was applied on a per unit basis to achieve the percent
reduction needed to meet the limit.  See below for some guidelines that
were used in applying controls.  

The emissions after control were calculated for each unit using the
following equation:

(‘02 Unit Emissions in tons) x (1 - % Control) = Actual Unit Emissions
After Control in Tons

The facility total emissions after control was calculated by summing the
emissions after control for all units at the facility.

The facility total emission reduction was calculated using the following
equation:

(’02 Facility Total Emissions in Tons) – (Facility Total Emissions
in Tons After Control) = Facility Emission Reduction in Tons 

For facilities with multiple units, several different combinations of
controls across the units often had to be tried before a level of
control that met the limits was achieved.  The control options that are
available vary depending on the cleaning machine type, the solvent, and
the percent control that is required.  Therefore, to aid in the
assigning of controls to specific units for the main population we
developed a table of the control options applicable to each cleaning
machine type, solvent, and the percent control that is required (Table
9).  

TABLE 9 - CONTROL OPTIONS FOR MAIN POPULATION

Cleaning Machine Type	Solvent	Percent Control Required to Meet Limit
Control Option

OTVC	PCE	78%-99%	PCE to MC

Vacuum



51%-77%	PCE to TCE

PCE to MC

Vacuum



31%-50%	Retrofit – 1.5FBR, WC, FRD



≤ 30%	Retrofit – 1.5FBR

Retrofit – CAD

	TCE	71%-99%	Vacuum



51% - 70%	TCE to MC

Vacuum



31%-50%	Retrofit – 1.5FBR, WC, FRD



≤ 30%	Retrofit – 1.5FBR

Retrofit – CAD

	MC	51%-99%	Vacuum



31%-50%	Retrofit – 1.5FBR, WC, FRD



≤ 30%	Retrofit – 1.5FBR

Retrofit – CAD

In-line/Cold	PCE	78%-99%	PCE to MC

Vacuum



31%-77%	PCE to TCE

PCE to MC

Vacuum



≤ 30%	Retrofit – 1.5FBR

Retrofit – CAD

	TCE	71%-99%	Vacuum



31%-70%	TCE to MC

Vacuum



≤ 30%	Retrofit – 1.5FBR

Retrofit – CAD

	MC	78%-99%	Vacuum



31% to 77%	TCE to MC

Vacuum



≤ 30%	Retrofit – 1.5FBR

Retrofit – CAD

	MC	0% to 99%	Vacuum

 

In cases where more than one control option is available, we made a
starting assumption regarding the distribution of controls for the main
population.  For example, we received comments on the proposed rule that
indicated that no more that 15% of the affected units should be assumed
to choose the solvent switching option.  Therefore, the solvent
switching option was not chosen for more than 15% of the affected units.
 In addition, it was assumed that approximately a third of the affected
units would choose to replace their machines with vacuum cleaning
machines.  Beyond these assumptions the assigned controls were the
product of the available controls and the emission reduction required to
achieve the limit being evaluated.   Table 10 shows the controls applied
for each of the limits evaluated for the main population and Table 11
shows the controls applied for the aerospace facilities.  The control
options used for web cleaning machines and tube manufacturers were
provided by industry.  For web cleaning machines, the only control
option applied was an automated gate, lengthening the cleaning process
area, providing an additional enclosure around all distillation units,
and modifying the capacity and process structure of the CADs.  For tube
manufacturers, the only control option applied was a side chamber.   For
Anniston Army Depot, a 1.5 freeboard ratio, working mode cover, and
freeboard refrigeration device were applied for the purpose of
estimating costs.  

Table 10 - Controls Applied for Main 2002 Population Excluding Tube,
Web,

Aerospace, and Anniston Army Depot

Applied Control Option	% of Units Assigned Control to Meet 25,000 kg MC
EQ Limit	% of Units Assigned Control to Meet 40,000 kg MC EQ Limit	% of
Units Assigned Control to Meet 60,000 kg MC EQ Limit	% of Units Assigned
Control to Meet 100,000 kg MC EQ Limit

Vacuum to Vacuum	38%	31%	36%	33%

CAD 	11%	15%	0%	0%

PCE to MC	4%	3%	4%	3%

PCE to TCE	5%	3%	4%	3%

TCE to MC	5%	11%	9%	9%

RETRO - 1.5FBR, WC, FRD	23%	23%	24%	36%

RETRO - 1.5FBR	15%	14%	24%	15%





Table 11- Controls Applied for Aerospace Facilities

Applied Control Option	% of Units Assigned Control to Meet 100,000 kg MC
EQ Limit	% of Units Assigned Control to Meet 250,000 kg MC EQ Limit

Vacuum Machine - Small - Baskets	8%	0%

Vacuum Machine - Small - Hanging	8%	0%

Vacuum Machine - Large - Baskets	23%	8%

Vacuum Machine - Large - Hanging	23%	0%

Enclosure, parts transporter, CAD, and distillation	15%	31%

Dual-Split Working-Mode Cover and Dual-Bank Freeboard Chillers	15%	31%

Enclosure	8%	8%



Below is an example of how controls were applied for an example
facility:

The facility has three units (Table 12), one using MC and two using PCE.
 The actual emissions were weighted to obtain the MC Equivalent
emissions for the facility.  Based on the MC equivalent emissions, the
facility must achieve an 88% reduction to meet a 60,000 kg MC equivalent
limit.  

Not adding any additional control to units one and three and replacing
unit two with a vacuum cleaning machine yields a reduction of 93%
[(40,700 kg –  2,700 kg)/40,700 kg = 93%].   

Table 12 - Example Facility Emissions and Controls

Unit	Solvent	Cleaner

Type	Emissions (kg)	MC Equivalent

Emissions (kg)	Controls Applied	Control Achieved	MC EQ  Emissions After
Control (kg)	Actual Emissions After Control (kg)

1	MC	Batch Cold	200	200	None	0%	200	200

2	PCE	Open Top	40,000	502,000	Vacuum to Vacuum	95%	25,100	2,000

3	PCE	Open Top	500	6,275	None	0%	6,275	500

Total	_-	-	40,700	508,475	-	96%	31,575	2,700



Below is an example of how the control costs were applied for the same
facility:

The individual control costs from Table 3 were applied to each control
device.  The costs associated with the replacement of the existing
equipment for unit 2 with a vacuum to vacuum cleaning machine are shown
in Table 13.  Solvent savings represents the cost savings that result
from reduced solvent purchases.  The solvent savings were calculated
using the unweighted emission reduction for the unit and the cost of
solvent per gallon as shown below:

(40,000 kg – 2,000 kg) x $2.32/kg = $88,160

Due to the solvent savings, the facility is projected to have a net cost
savings of $31,665/year.

Table 13 - Example Facility Costs

Unit	Solvent	Control Achieved	Total Capital Costs Per Unit	Annualized
Capital Costs Per Unit	O&M Costs Per Unit	Solvent Savings Per Unit	Total
Annualized Costs Per Unit	Controls Applied

1	MC	0%	0	0	0	0	0	None

2	PCE	95%	$399,000	$37,663	$18,832	($88,160)	($31,665)	Vacuum to Vacuum

3	PCE	0%	0	0	0	0	0	None

Total	-	93%	$399,000	$37,663	$18,832	($88,160)	($31,665)	-



	Because only a finite number of control options are available, in some
cases, the emission reduction required for a facility to meet the
control option limits was exceeded.  In other words, if a 50% emission
reduction was required, the controls applied achieved a reduction
greater than 50%.  In some cases, the emission reduction achieved was a
little less than that required for a facility to meet the control option
limits.  For example, a facility that required a 99% emission reduction
to meet the control option limits, could at best achieve a 95% emission
reduction with the installation of a vacuum to vacuum cleaning machine. 


7.0 Summary of Cost and Emission Reduction Results

	The costs and emission reductions for all units at all facilities with
emissions above the control option limits were totaled to yield the
total national costs and emission reductions.  Table 14 summarizes the
national capital costs, national annualized capital costs, national O&M
Costs, national solvent savings, and the total national annual emission
control costs for each control option.  

	Table 14 shows that control costs increase and solvent savings increase
as the emission limit is set lower.  The lower the limit is established,
the greater the number of units that must be controlled to achieve the
limit.  Emission reductions are greater the lower the limit is
established, therefore, the solvent savings are greater.  

	For the main population total annual emission control costs range from
a savings of $1.3 million/year for the 100,000 kg and the 60,000 kg MC
equivalent control options to a cost of $1.2 million/year for the 25,000
kg MC control option.  Capital costs for the 4 control options range
from approximately $8.5 million for the 100,000 kg MC equivalent option
to $41.5 million for the 25,000 kg MC equivalent option.  Annualized
capital costs range from $800,000/year for the 100,000 kg MC equivalent
option to $4 million/year for the 25,000 kg MC equivalent option. 
Operating and maintenance costs range from $417,000 for the 100,000 kg
MC equivalent option to $2.2 million for the 25,000 kg MC equivalent
option.  Solvent savings have a significant impact on total annual
costs, ranging from a savings of over $2.5 million/year for the 100,000
kg MC option to a savings of over $5 million/year for the 25,000 kg MC
equivalent option.  As indicated previously, solvent savings represent
the cost savings that result from reduced solvent purchases.   The cost
effectiveness values for the 100,000 kg MC option and the 60,000 kg MC
option indicate cost savings at -$1,146/ton and -$832/ton, respectively.
 The cost effectiveness values for the 40,000 kg MC option and the
25,000 kg MC option indicate costs at $74/ton and $520/ton,
respectively.

	For the aerospace population total annual emission control costs range
from $115,000/year for the 250,000 kg to a cost of $627,000/year for the
100,000 kg MC control option.  Capital costs range from approximately
$3.2 million for the 250,000 kg MC equivalent option to $9 million for
the 100,000 kg MC equivalent option.  Annualized capital costs range
from $332,000/year for the 250,000 kg MC equivalent option to
$870,000/year for the 100,000 kg MC equivalent option.  Operating and
maintenance costs range from $166,000 for the 250,000 kg MC equivalent
option to $435,000 for the 100,000 kg MC equivalent option.  Solvent
savings have a significant impact on total annual costs, ranging from a
savings of over $383,000/year for the 250,000 kg MC option to a savings
of $627,000/year for the 100,000 kg MC equivalent option.  The cost
effectiveness values for the aerospace options are $627/ton and
$1,933/ton for the 250,000 kg MC option and the 100,000 kg MC option,
respectively.

	Costs for Anniston Army Depot at the 100,000 kg MC option result in a
net annualized cost savings of $56,000/year.  This is due to the solvent
cost savings of $157,000/year that are expected to result.  These
solvent cost savings more than offset the annualized capital costs
($59,000/year) and O&M costs ($42,000).  The cost effectiveness of
control at the 100,000 kg MC level is -$625/year for the Anniston Army
Depot.  

	Costs for facilities performing web cleaning to achieve an 80% overall
control efficiency and for tube manufacturers to achieve an additional
10% additional emission reduction high relative to the costs for the
main population, Aerospace facilities, and Anniston Army Depot.  Capital
costs are $5.7 million and $6.6 million for web and tube, respectively. 
Solvent savings are $345,000/year for web and $400,000 for tube.  The
cost effectiveness of control for the web and tube industry sectors is
$3,421/year and $3,624/year, respectively.  

 	For additional details regarding the development of the costs and
emission reduction impact for the final rule refer to the cost
spreadsheet.  These spreadsheets also contain additional summary tables
not reproduced here.  

Table 14 – Summary of National Costs and Emission Reductions for
Options Evaluated for Final Rule

Description	Options in kg Methylene Chloride Equivalents for the
Population without Tube, Web, Aerospace, and Anniston Army Depot	Options
in kg Methylene Chloride Equivalents for  Aerospace 	Options in kg
Methylene Chloride Equivalents for Anniston Army Depot	Web Cleaning	Tube
Manufacturing

	100,000	60,000	40,000	25,000	100,000	250,000	100,000	80% Overall
Efficiency	10% Emission Reduction

1.  National Total Capital Costs	$8,473,912	$15,144,335	$26,107,974
$41,560,768	$9,031,798	$3,168,991	$536,431	$5,739,211	$6,623,853

	 

	 	 

 	 	 

2.  National Annualized Total Capital Costs	$812,736	$1,448,795
$2,557,105	$4,028,968	$870,590	$332,258	$59,007	$631,313	$728,624

	 

	 	 	 	 	 	 

3.  National Total Annual Operation and Maintenance Costs	$416,744
$724,063	$1,530,058	$2,248,569	$435,295	$166,129	$42,143	$315,657
$364,312

	 

	 	 	 	 	 	 

4.  National Emission Reduction (true tons)	1,104	1,594	1,759	2,351	324
183	89	176	192

	 

	 	 	 	 	 	 

5.  National Solvent Credit	-$2,493,972	-$3,498,691	-$3,957,103
-$5,054,032	-$679,102	-$383,369	-$156,912	-$345,068	-$396,125

	 

	 	 	 	 	 	 

6.  National Net Annualized Cost of Control ($/yr)

[Annualized Total Capital Costs + O&M Costs + Solvent Credit]
-$1,264,492	-$1,325,832	$130,060	$1,223,505	$626,784	$115,018	-$55,761
$601,902	$696,810

	 

	 	 	 	 	 	 

7.  National Cost Effectiveness ($/ton)

[Net Annualized Cost of Control/Emission Reduction]	-$1,146	-$832	$74
$520	$1,933	$627	-$625	$3,421	$3,624



Appendix

Table A-1.  Basis of Costs for 1.50 Freeboard Ratio, Freeboard
Refrigeration Device, Working Mode Cover, and Internal Hoist





*Average Size*

1.  Capital Costs	1 m2	4m2	2.5 m2

	2005 Installed Capital Costs [1]	$18,000	$30,000	$24,000.00

	Cost of Additional Floor Space[2]	$290	$3,000	$1,645.15







	Total Capital Costs	$18,290	$33,000	$25,645.15







	Annualized Total Capital Costs (15yr @ 7%)	$2,012	$3,630	$2,820.97







2. Annual Operation Costs





A.  Labor





   -1.50 Freeboard Ratio	$0	$0	$0.00

	   -Working Cover	$0	$0	$0.00

	   -Freeboard Refrigeration Device [3]	$146	$146	$146.00

	   TOTAL LABOR	$146	$146	$146.00







	B.  Energy [4, 5]





   -1.50 Freeboard Ratio	$0	$0	$0.00

	   -Working Cover	$252	$252	$252.00

	   -Freeboard Refrigeration Device	$211	$971	$590.95

	   TOTAL ENERGY	$463	$1,222	$842.95







	C.  Misc. Operating Costs [6]	$732	$1,320	$1,025.81







	Total Annual Operation Costs	$1,340	$2,688	$2,014.75







3.  Total Annualized Cost (Annualized Total Capital Cost plus Total
Annual Operating Costs)	$3,352	$6,318	$4,835.72

	[1] - Telecommunication J. Goldsmith, e2M with Carl Wolf, Ultronix. 
September 14, 2005.

	[2] - Scaled from 1990 using the PPI.

	[3] - Assumed operating 8,760 hrs per year.  1,095 shifts/year *
$13.29/hr * 0.01 hrs/shift = labor cost.  Labor rate was obtained from
BLS - Private Industry Blue Collar Machine Operators, July 2003.

	[4] - For 1m2 size WC uses 0.56 kWh and FRD uses 0.47 kWh.  For 4m2
size, WC uses 0.56 kWh and FRD uses 2.16 kWh.  Source: IV-B-04  Assumed
devices are operating 8,760 hours per year.   

	[5] Source: Table 7.4 - Average Retail Price of Electricity to Ultimate
Customers by End-Use Section,  1993 through 2003.  5.13 cents per kwh
represents the 2003 Industrial value. 

	[6] - Calculated as 4% of capital costs (Source 1).



Table A-2.  Basis of Costs for 1.5 Freeboard Ratio





*Average Size*

1.  Capital Costs	1 m2	4 m2	2.5 m2

	2005 Installed Capital Costs [1, 2]	$8,152	$32,608	$20,380.00 





 

	Total Capital Costs	$8,152	$32,608	$20,380.00 





 

	Annualized Total Capital Costs (15yr @ 7%)	$897	$3,587	$2,241.80 





 

2. Annual Operation Costs

	 

	A.  Labor 	$0	$0	$0.00 

	B.  Energy	$0	$0	$0.00 

	C.  Misc. Operating Costs	$0	$0	$0.00 





 

	Total Annual Operation Costs	$0	$0	$0.00 





 

3.  Total Annualized Cost (Annualized Total Capital Cost plus Total
Annual Operating Costs)	$897	$3,587	$2,241.80 













	[1] - Telecommunication.  J. Goldsmith of e2M with Rod Murphy of
Degreasing Devices.  September 15, 2005. 

	[2] - Based on an Open Top Vapor Cleaner Size of 4.6 m2.  

	[3] - Conservatively used the cost for PCE, which is cheaper than TCE.



Table A-3.  Capital Costs for Vacuum-to-Vacuum Cleaning Machines

&

2

3

J

˜

Ç

hh 

hh 

8

ˆ

8

8

ˆ

8

8

ˆ

8

8

ˆ

8

8

ˆ

8

8

ˆ

8

8

ˆ

8

8

ˆ

8

hh 

hh 

 hh 

hh 

  hh 

hh 

gdh 

8

ˆ

8

葠ː摧㉖s

heI

 heI

heI

 heI

heI

heI

heI

혈F鐃᳿ꐋⰖ"蠆

蠆

蠆

摧䥥

摧䥥

摧䥥

혈F鐃᳿ꐋⰖ"蠆

蠆

蠆

葞и摧䶦t

摧䥥

hp

gdŸ_

葞ː葠ː摧抠ò

葞֠摧嵙ô

P	¤

P	¤

P	¤

P	¤

P	¤

P	¤

P	¤

P	¤

P	¤

P	¤

P	¤

P	¤

P	¤

P	¤

摧徟

摧徟

葞Ũ摧੧

h]

Ff

kd

gdp

ŔȀ沖܀ດࠁ峖Ѐ”ÿ

ƒ 

¡

ƒ 

c - Based on Costs from Tiyoda-Serec Corporation and Baron Blakeslee.

	d - This was calculated using the relationship established between the
solvent-air interface of an open-top and its cleaning capacity that was
established during the development of the NESHAP (A-92-39, IV-B-2).  The
following equation was developed from the NESHAP analysis:

Solvent-Air Interface in square meters = 2.2 (cleaning volume in cubic
meters)0.6



 PAGE   

 PAGE   25 

