United States
Environmental Protection
Agency
Office of Air Quality
• Planning and Standards
Research Triangle Park NC 27711
EPA-453/D-93-016
October 1992
Air
$ EPA Ethylene Oxide
Emissions from
Commercial
Sterilization/Fumigation
Operations
Background
Information for
Proposed Standards
DRAFT EIS
-------
-------
ETHYLENE OXIDE EMISSIONS FROM COMMERCIAL
STERILIZATION/FUMIGATION OPERATIONS --
BACKGROUND INFORMATION FOR PROPOSED
STANDARDS
Final
-------
ETHYLENE OXIDE EMISSIONS FROM COMMERCIAL STERILIZATION/FUMIGATION
OPERATIONS — BACKGROUND INFORMATION FOR PROPOSED STANDARDS
Emission Standards Division
U. S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
March 1993
11
-------
This report has been reviewed by the Emission Standards
Division of the Office of Air Quality Planning and Standards, EPA
and approved for publication. Mention of trade names or
commercial products is not intended to constitute endorsement or
recommendation for use. Copies of this report are available
through the Library Services Office (MD-35), U. S. Environmental
Protection Agency, Research Triangle Park, North Carolina 27711;
from the Office of Air Quality Planning and Standards Technology
Transfer Network, U. S. Environmental Protection Agency, Research
Triangle park, North Carolina 27711; or, for a fee, from the
National Technical Information Services, 5285 Port Royal Road,
Springfield, Virginia 22161.
Publication No. EPA-435/D-93-016
111
-------
ENVIRONMENTAL PROTECTION AGENCY
Background Information
and Draft
Environmental Impact Statement
or Commercial Sterilization/Fumigation Operations
Prepared by:
Bruce Jordan-
Director./Emission standards Division
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
j^Date)
27711
The proposed national emission standard would limit emissions
of ethylene oxide from existing and new commercial
sterilization/fumigation operations. The proposed standards
implement Section 112 of the Clean Air Act as amended in 1990
and are based on the Administrator's determination of
July 16, 1992 (57 FR 31576) that commercial sterilization
sources generate a large amount of ethylene oxide, a
hazardous air pollutant listed in Section 112(b) of;the Act.
Copies of this document have been sent to the.following
Federal Departments: Labor, Health and Human Services,
Defense, Transportation, Agriculture, Commerce, Interior, and
Energy; the National Science Foundation; the Council on
Environmental Quality; members of the State and Territorial
Air Pollution Program Administrators; the Association of
Local Air Pollution Control Officials; EPA Regional
Administrators; Office of Management and Budget; and other
interested parties.
The comment period for review of this document is 60 days.
Mr. David Markwordt, Chemicals and Petroleum Branch;
telephone (919) 541-0837, may be contacted regarding the date
of the comment period.
For additional information contact:
Mr. David Markwordt (MD-13)
Chemicals and Petroleum Branch
U. S. Environmental Protection Agency
Research Triangle Park, N. C. 27711
Telephone: (919) 541-0837
IV
-------
5. Copies of this document may be obtained from:
U. S. EPA Library (MD-36)
Research Triangle Park, N.C.
27711
National Technical Information Service
5285 Port Royal Road
Springfield, Virginia 22161
-------
TABLE OF CONTENTS
LIST OF FIGURES
LIST OF TABLES
CHAPTER 2.
CHAPTER 3.
JRES
LES .
SUMMARY .
1.1 STATUTORY AUTHORITY ...
1.2 REGULATORY ALTERNATIVES
1.3 ENVIRONMENTAL IMPACT
1.4 COST IMPACT
1.5 ECONOMIC IMPACT
INTRODUCTION
2.1 BACKGROUND AND AUTHORITY FOR STANDARDS . .
2.2 SELECTION OF POLLUTANTS AND
SOURCE CATEGORIES .
2.3 PROCEDURE FOR DEVELOPMENT OF NESHAP . . .
2.4 CONSIDERATION OF COSTS
2.5 CONSIDERATION OF ENVIRONMENTAL IMPACTS . .
2.6 RESIDUAL RISK STANDARDS
ETHYLENE OXIDE STERILIZATION/ FUMIGATION
PROCESSES AND EMISSIONS
3.1 BACKGROUND INFORMATION .....
3.2 PROCESS DESCRIPTION
3.2.1 Bulk Sterilization .
3.2.2 Single-Item Sterilization System .
3.2.3 Spice Fumigators
3.2.4 Library and Museum Fumigators . . .
3.2.5 Beehive Fumigators
3 . 3 EMISSION SOURCES . .
3.3.1 Sterilization Chamber Vents ....
3.3.2 Sterilization Chamber Vacuum
Pump Drains
3.3.3 Chamber Exhaust Vent .
3.3.4 Aeration Room Vent
3.3.5 Equipment Leaks .
3.3.6 Storage and Handling .
3.4 EMISSION ESTIMATES
3.4.1 Commercial Sterilization
Facilities
3.5 CURRENT REGULATIONS
3.5.1 Occupational Safety and Health
Administration Standard . . . . . .
3.5.2 State Regulations .
3 . 6 REFERENCES FOR CHAPTER 3 .
Page
X
xi
1-1
1-1
1-1
1-2
1-2
1-2
2-1
2-1
2-5
2-6
2-9
2-10
2-11
3-1
3-1
3-4
3-4
3-16
3-18
3-18
3-19
3-19
3-21
3-21
3-21
3-23
3-23
3-23
3-23
3-23
3-27
3-27
3-27
3-29
VI
-------
TABLE OF CONTENTS (continued)
CHAPTER 4. EMISSION CONTROL TECHNIQUES 4-1
4.1 BULK STERILIZATION PROCESSES 4-1
4.1.1 Sterilization Chamber Vent
Emissions ............. 4-1
4.1.2 Sterilization Chamber Vacuum Pump
Drain Emissions 4-16
4.1.3 Chamber Exhaust Emissions . . . . . 4-18
4.1.4 Aeration Room Vent Emissions . . . 4-20
4.2 OTHER STERILIZATION PROCESSES 4-25
4.2.1 Single-Item Sterilization 4-25
4.2.2 Fumigation with Portable Units . . 4-25
4.3 ALTERNATIVES TO EO STERILIZATION 4-26
4.4 RETROFIT CONSIDERATIONS 4-26
4.5 IMPACTS OF CFC REGULATION ON EO
EMISSION CONTROLS 4-27
4.6 REFERENCES FOR CHAPTER 4 4-27
CHAPTER 5. REGULATORY ALTERNATIVES 5-1
5.1 INTRODUCTION 5_!
5.2 REGULATORY ALTERNATIVES ... 5-2
CHAPTER 6. ENVIRONMENTAL IMPACTS 6-1
6.1 AIR POLLUTION IMPACTS 6-1
6.1.1 Baseline Emissions and
Emission Reduction 6-1
6.1.2 Secondary Impacts 6-1
6.2 WATER QUALITY IMPACTS 6-3
6.3 SOLID WASTE IMPACTS 6-4
6.4 ENERGY IMPACTS 6-4
6.5 OTHER ENVIRONMENTAL CONCERNS 6-6
6.5.1 Irreversible and Irretrievable
Commitment of Resources 6-6
6.5.2 Environmental Impact of Delayed
Standards 6-6
6.6 REFERENCES FOR CHAPTER 6 6-6
CHAPTER 7. EMISSION CONTROL COSTS 7-1
7.1 STERILIZATION VENT CONTROL COSTS 7-2
7.1.1 Description of Components Costed . 7-3
7.1.2 General Assumptions 7-3
7.2 CHAMBER EXHAUST CONTROL COSTS ...... 7-5
7.2.1 Description of Components Costed . 7-5
7.2.2 General Assumptions 7-6
vii
-------
TABLE OF CONTENTS (continued)
Page
7.3 AERATION ROOM CONTROL COSTS ....... 7-6
7.3.1 Description of Components Costed . 7-7
7.3.2 General Assumptions 7-7
7.4 RESULTS OF COST ANALYSIS . 7-9
7.5 OTHER COST CONSIDERATIONS 7-9
7.6 REFERENCES FOR CHAPTER 7 7-12
CHAPTER 8. THE ECONOMIC IMPACTS OF THE CANDIDATE
NESHAP CONTROLS 8-1
8.1 INTRODUCTION .. ; . . 8-1
8.2 ETHYLENE OXIDE STERILIZATION . . 8-2
8.2.1 Process Inputs 8-2
8.3 SUBSTITUTION POSSIBILITIES AND THE
PRICE ELASTICITY OF DEMAND 8-3
8.4 SUPPLY OF EO STERILIZATION SERVICES . . , 8-4
8.4.1 National Summary of Ethylene
Oxide Sterilization 8-5
8.4.2 Industry Groups Supplying EO
Sterilization Services 8-7
8.5 DEMAND FOR ETHYLENE OXIDE
STERILIZATION SERVICES ... 8-34
8.6 ECONOMIC EFFECTS OF CANDIDATE NESHAP
CONTROLS UNDER THREE CONTROL OPTIONS . . . 8-35
8.6.1 The Three Control Options ..... 8-35
8.6.2 Theoretical Framework for
Economic Impact Analysis 8-36
8.6.3 Analytical Procedure . 8-42
8.6.4 Results 8-48
8.7 EFFECTS OF THE REGULATION ON SMALL
BUSINESSES • . . 8-59
8.7.1 Requirements of the Regulatory
Flexibility Act 8-59
8.7.2 Small Businesses Performing
Ethylene Oxide Sterilization . . . 8-61
8.7.3 Small Businesses in the Contract
Sterilizer Industry Group . . . . . 8-62
8.7.4 Substitution of Contract
Sterilization for In-House
Sterilization 8-63
8.7.5 Small Business Impacts in Other
Industry Groups 8-70
8.7.6 Summary 8-72
8.8 CONCLUSIONS 8-72
8.8.1 Effects on Existing Facilities . . 8-72
8.8.2 Effects on New Facilities 8-77
8.9 REFERENCES FOR CHAPTER 8 . . 8-78
viii
-------
TABLE OF CONTENTS (continued)
APPENDIX A EVOLUTION OF THE BACKGROUND INFORMATION
DOCUMENT
APPENDIX B INDEX TO ENVIRONMENTAL CONSIDERATIONS ....
APPENDIX C EMISSION SOURCE TEST DATA . .
APPENDIX D EMISSION MEASUREMENT AND CONTINUOUS
MONITORING
APPENDIX E SUPPLEMENTAL INFORMATION FOR THE COST
ANALYSIS
Page
A-l
B-l
C-l
D-l
E-l
IX
-------
LIST OF FIGURES
Page
Figure 3-1. Schematic of a gas sterilizer . . 3-5
Figure 3-2. Sterilization cycle for 12/88 . 3-12
Figure 3-3. Sterilization cycle for pure EO 3-13
Figure 3-4. Schematic of emission sources at commercial
sterilization facilities . 3-20
Figure 3-5. Hydrolysis rates of dilute, neutral aqueous
solutions of ethylene oxide . 3-22
Figure 4-1. Countercurrent packed scrubbing system .... 4-5
Figure 4-2. Detoxification tower control system ..... 4-7
Figure 4-3. Catalytic oxidation system . 4-11
Figure 4-4. Condensation/reclamation system , . . 4-14
Figure 4-5a. Once-through liquid-ring vacuum pump 4-17
Figure 4-5b. Recirculating liquid-ring vacuum pump .... 4-17
Figure 8-1. Demand curve for Commodity Q . 8-37
Figure 8-2. Supply curve for Commodity Q . 8-38
Figure 8-3. Market equilibrium with and without an
upward shift in the supply curve due to
ethylene oxide emission controls . 8-40
Figure 8-4. The market for contract sterilization without
the air emission standard in place ...... 8-64
Figure 8-5. Marginal cost curves for a contract
sterilizer and an in-house sterilizer,
with and without the air emission standard
in effect 8-66
Figure 8-6. The market for contract sterilization with
the air emission standard in effect 8-69
Figure C-l. Sampling point locations . c-9
-------
LIST OF TABLES
TABLE 1-1. POTENTIAL NATIONWIDE PERCENT EMISSION
REDUCTION AND NATIONWIDE AIR, WASTEWATER,
SOLID WASTE, AND ENERGY IMPACTS 1-3
TABLE 1-2. NATIONWIDE REGULATORY ALTERNATIVE COST
IMPACTS 1-4
TABLE 3-1. LOCATIONS OF FACILITIES—EPA COMMERCIAL
STERILIZATION DATA BASE 3-3
TABLE 3-2. NUMBER OF FACILITIES AND STANDARD INDUSTRIAL
CLASSIFICATION (SIC) PER INDUSTRY CATEGORY—
EPA COMMERCIAL STERILIZATION DATA BASE .... 3-5
TABLE 3-3. CHAMBER SIZES—EPA COMMERCIAL STERILIZATION
DATA BASE 3-7
TABLE 3-4. PHYSICAL AND CHEMICAL PROPERTIES OF ETHYLENE
OXIDE, DICHLORODIFLUOROMETHANE,
AND CARBON DIOXIDE 3-9
TABLE 3-5. STERILANT GAS TYPE USAGE—EPA COMMERCIAL
STERILIZATION DATA BASE 3-10
TABLE 3-6. AVERAGE EMISSIONS FROM COMMERCIAL
STERILIZATION FACILITIES—EPA COMMERCIAL
STERILIZATION DATA BASE . 3-25
TABLE 3-7. STATE REGULATIONS FOR ETHYLENE OXIDE
EMISSIONS 3-28
TABLE 4-1*. ETHYLENE OXIDE EMISSION CONTROL DEVICES FOR
STERILIZER VENTS—EPA COMMERCIAL
STERILIZATION DATA BASE 4-2
TABLE 5-1. REGULATORY ALTERNATIVES 5-3
TABLE 6-1. NATIONWIDE AIR IMPACTS 6-2
TABLE 6-2. POTENTIAL NATIONWIDE WASTEWATER, SOLID
WASTE, AND ENERGY IMPACTS 6-5
TABLE 7-1. NATIONWIDE REGULATORY ALTERNATIVE COST
IMPACTS 7-10
TABLE 7-2. REPRESENTATIVE FACILITY COST IMPACTS 7-11
XI
-------
LIST OF TABLES (continued)
TABLE 8-1. SUMMARY STATISTICS ON THE USE OF STERILANT
GAS AT 188 COMMERCIAL STERILIZATION
FACILITIES 8_6
TABLE 8-2. STANDARD INDUSTRIAL CLASSIFICATION CODES
FOR 188 COMMERCIAL STERILIZATION FACILITIES . 8-8
TABLE 8-3. RECENT PERFORMANCE AND FORECAST DATA FOR
MEDICAL DEVICE SUPPLIERS (SIC 3841 AND 3842) . 8-11
TABLE 8-^4. SUMMARY STATISTICS ON STERILIZATION CHAMBERS
AND GASES USED BY 62 MEDICAL DEVICE
SUPPLIERS 8-13
TABLE 8-5. SUMMARY STATISTICS ON STERILIZATION CHAMBERS
AND GASES USED BY 24 OTHER HEALTH-RELATED
SUPPLIERS 8_15
TABLE 8-6. RECENT PERFORMANCE AND FORECAST DATA FOR
PHARMACEUTICAL MANUFACTURERS (SIC 2834) . . . 8-17
TABLE 8-7. SUMMARY STATISTICS ON STERILIZATION CHAMBERS
AND GASES USED BY 39 PHARMACEUTICAL
MANUFACTURERS 8-19
TABLE 8-8. RECENT PERFORMANCE DATA FOR SPICE
MANUFACTURERS (SIC 2099) s-21
TABLE 8-9. 1987 PERFORMANCE DATA FOR SPICE
MANUFACTURERS (SIC 2099) 8-23
TABLE 8-10. SUMMARY STATISTICS ON STERILIZATION CHAMBERS
AND GASES USED BY 23 SPICE MANUFACTURERS . . . 8-24
TABLE 8-11. SUMMARY STATISTICS ON STERILIZATION CHAMBERS
AND GASES USED BY 13 MUSEUMS AND
LIBRARIES 8-27
TABLE 8-12. SUMMARY STATISTICS ON STERILIZATION
CHAMBERS AND GASES USED BY 10 LABORATORIES . . 8-30
TABLE 8-13. SUMMARY STATISTICS ON STERILIZATION
CHAMBERS AND GASES USED BY 17 CONTRACT
STERILIZERS ........ 8-33
TABLE 8-14. CUMULATIVE TOTAL ANNUAL COMPLIANCE COST (TAG)
UNDER THE THREE CONTROL OPTIONS, FOR AFFECTED
INDUSTRY GROUPS 8-50
XXI
-------
TABLE 8-15.
LIST OF TABLES (continued)
COUNT OF FACILITIES HAVING POSITIVE AND
ZERO CUMULATIVE TOTAL ANNUAL COMPLIANCE
COST (TAC) UNDER THE THREE CONTROL
OPTIONS, FOR AFFECTED INDUSTRY GROUPS ,
Page
8-51
TABLE 8-16. CUMULATIVE TOTAL ANNUAL COMPLIANCE COST
(TAC) PER CUBIC METER OF FACILITY CHAMBER
VOLUME, FOR AFFECTED INDUSTRY GROUPS 8-53
TABLE 8-17. CUMULATIVE TOTAL ANNUAL COMPLIANCE COST
(TAC) PER METRIC TON OF ETHYLENE OXIDE USED
BY FACILITY, FOR AFFECTED INDUSTRY GROUPS . . 8-55
TABLE 8-18. CUMULATIVE TOTAL ANNUAL COMPLIANCE COST
(TAC) AS A PERCENTAGE OF BASELINE ANNUAL
STERILIZATION COSTS, FOR AFFECTED INDUSTRY
GROUPS 8-57
TABLE 8-19. CUMULATIVE TOTAL ANNUAL COMPLIANCE COST
(TAC) AS A PERCENTAGE OF ANNUAL FACILITY
SALES, FOR AFFECTED INDUSTRY GROUPS 8-60
TABLE 8-20. SMALL BUSINESSES IN THE INDUSTRY GROUPS
PERFORMING IN-HOUSE STERILIZATION 8-71
TABLE A-l. EVOLUTION OF THE BACKGROUND
INFORMATION DOCUMENT A-l
TABLE C-l. SUMMARY OF TEST RESULTS C-2
TABLE C-2. SUMMARY OF FIELD TEST AT BURRON MEDICAL . . . C-5
TABLE C-3. SUMMARY OF EMISSION MEASUREMENTS AND CONTROL
EFFICIENCIES FOR EMPTY CHAMBER TESTS AT
BURRON MEDICAL C-7
TABLE C-4. SUMMARY OF EMISSION MEASUREMENTS AND CONTROL
EFFICIENCIES FOR TESTS AT THE MCCORMICK AND
COMPANY, INC., SPICE MILL C-12
TABLE C-5. SUMMARY OF EMISSION MEASUREMENTS AND CONTROL
EFFICIENCIES FOR EMPTY CHAMBER TESTS
AT CHESEBOROUGH PONDS C-l5
TABLE E-l. COST OF DAMAS SCRUBBER MODELS E-3
TABLE E-2. CAPITAL AND ANNUALIZED COSTS OF INSTALLING
SCRUBBERS E-4
Xlll
-------
LIST OF TABLES (continued)
TABLE E-3.
TABLE E-4.
TABLE E-5.
TABLE E-6.
TABLE E-7.
TABLE E-8.
TABLE E-9.
TABLE E-10.
TABLE E-ll.
TABLE E-12.
TABLE E-13.
TABLE E-14.
TABLE E-15.
TABLE E-16.
TABLE E-17.
DATA USED TO CALCULATE SCRUBBER EQUIPMENT
CAPITAL COSTS
CAPITAL COST OF CHECK VALVE FOR CHAMBER
DATA USED TO CALCULATE CONTROL DEVICE
ANNUALIZED COSTS
MISCELLANEOUS OPERATING COSTS
COST OF ETO ABATOR CATALYTIC OXIDERS
AERATION ROOM GAS/SOLID REACTANT CONTROL
COST ANALYSIS
AERATION ROOM CATALYTIC OXIDATION CONTROL
COST ANALYSIS
CAPITAL AND ANNUAL COSTS OF INSTALLING
SCRUBBERS
CAPITAL AND ANNUAL COSTS OF INSTALLING
SCRUBBERS TO CONTROL CHAMBER EXHAUST VENTS .
CAPITAL AND ANNUAL COSTS OF GAS/SOLID
REACTOR TO CONTROL AERATION UNITS AT AN
EXAMPLE FACILITY
CAPITAL AND ANNUALIZED COSTS OF CATALYTIC
OXIDATION AT AN EXAMPLE FACILITY
INCREMENTAL CAPITAL COSTS OF MANIFOLDING
STERILIZATION CHAMBERS ....
DUCTWORK COSTS OF MANIFOLDING CHAMBER EXHAUST
VENTS TO A SCRUBBER . . .
DUCTWORK COSTS OF MANIFOLDING AERATION
UNITS TO A GAS/SOLID REACTOR
CHEMICAL ENGINEERING COST INDICES
Page
E-5
E-6
E-7
E-8
E-9
E-18
E-24
E-32
E-34
E-36
E-38
E-42
E-44
E-4 5
E-48
XIV
-------
1.0 SUMMARY
1.1 STATUTORY AUTHORITY
National emission standards for hazardous air pollutants are
established in accordance with Section 112(b)(l)(B) of the Clean
Air Act (42 U.S.C. 7412), as amended. Emission standards under
Section 112 apply to new and existing sources of a substance that
has been listed as a hazardous air pollutant. This study
examines emissions of ethylene oxide (EO) from commercial
sterilization and fumigation industries.
1.2 REGULATORY ALTERNATIVES
Five regulatory alternatives representing selected
combinations of control options were developed to evaluate the
environmental and cost impacts of differing control strategies.
Regulatory Alternative A represents the maximum level of
control with 99 percent of the EO emissions from all emissions
points associated with commercial sterilization operations
captured and controlled. Regulatory Alternative B represents the
maximum level of control of all emissions points that exceed an
EO use cutoff. Regulatory Alternative C represents control of
the sterilizer vent, vacuum pump drain, and aeration room
emissions at this same level of EO use. Additional controls on
chamber exhaust emissions are not anticipated under this
alternative. Regulatory Alternative D represents control of only
the sterilizer vent and vacuum pump drain emissions for
facilities using 270 kilograms per year (kg/yr) (600 pounds per
year [lb/yr]) or more of EO. Regulatory Alternative E represents
control of these same two emissions points for facilities using
900 kg/yr (2,000 lb/yr) or more of EO. Regulatory Alternative E
1-1
-------
represents the maximum achievable control technology (MACT) floor
determination.
1.3 ENVIRONMENTAL IMPACT
Table 1-1 summarizes the environmental impacts of the
regulatory alternatives. At the MACT floor (Regulatory
Alternative E), the nationwide EO emissions are estimated at
120 megagrams per year (Mg/yr) (132 tons per year [tons/yr]).
The lower EO use cutoff proposed under Regulatory Alternative D
reduces the estimated nationwide EO emissions to 109 Mg/yr
(120 tons/yr). The control options under Regulatory Alternative
C reduce the nationwide emissions to 68 Mg/yr (75 tons/yr).
Under Regulatory Alternatives B and A, the control options
proposed would reduce the nationwide emissions of EO to 30 Mg/yr
and 11 Mg/yr (33 tons/yr and 12 tons/yr), respectively.
The potential impacts of these regulatory alternatives on
wastewater, solid waste, and energy are also shown in Table 1-1.
It is expected that the wastewater and solid waste impacts will
be insignificant because of the recycling of ethylene glycol and
reactant.
1.4 COST IMPACT
The nationwide cost impacts of the regulatory alternatives
are summarized in Table 1-2. The costs associated with
Regulatory Alternative E (MACT floor regulation) may require a
nationwide capital investment of about $3.8 million. The control
measures in Regulatory Alternatives D, C, B, and A may each
require a nationwide capital investment of about $4.3 million,
$6.4 million, $9.2 million, and $12 million respectively. These
cost figures were determined using fourth quarter 1987 dollars.
1.5 ECONOMIC IMPACT
The economic impacts associated with the regulation of
commercial sterilizers are not significant. In general,
sterilization costs represent a small fraction of total
production costs for facilities in industries in which
sterilization is not the main source of revenue. Thus, any cost
increases caused by the regulation will not significantly
increase total production costs. Furthermore, because total
1-2
-------
Q?
IH
W
Q
H
0
H
1
Q
EH IK**
o g
Q ^
*§
CO §
o
H Q
co 9
co s
H *"
W y"
EH EH
W<
Jrf
W Q
CU jj
§8
H
1"
Si
r^J ja
SEH
rrt
<1
H •*
EH
H
O
O<
€
H
1
rH
03
§*?-§?
C 'n -H 0
C-H ** °
IH g ~_~
rH "H 3
JS 8 g*
o *t1 o s
S ° OS
Total annual
solid waste
production,
Mg/(tons)c'd
rH "g
3 0) >Q
C -P C *
C (0 Off)
0) 4J r7
rH -P O (0 3 17
4J «j -O —
0 5 O
EH M
ft
C X«4
O >i
•H — .
tO £ (-1
•H >i
g C ^
Q) O tO
•H C
>H 4J O
•H O 4->
«J 3 s-'
O 0)
W i*
Q)
73 C *
•HOC
>-H O
C tO-H
O W -P
.H >H U
4-> g 3
flj 0> 73
21 0)
*H
r7>
(d r-*
rH g
O
O
VO
^
en
o
CM
O
ON
rH
O
O
O
cn
VO
in
o
in
^
CM
O
rH
^
H
VO
0
rH
a\
o\
rt
o
o
^
CM
in
rH
n
rH
O
O
O
VO
VO
in
o
^
CM
CO
rH
^
CM
'S'
O
H
^
en
CO
o
o
%.
CM
in
rH
H
o"
O
0
VO
VO
in
o
^
CM
r»
o
rH
*
O
O
rH
•*
cn
O
0
o
o"
o
o
VO
VO
in
o
^
CM
CM
VO
0
*
n
VO
en
o
Cf>
Q
0
o
o
o
o
rH
VO
in
o
CM
hi.
CM
a\
0
••
CM
in
cn
cn
CO
H
O
O
• ^»l
iH rH
O tJ*
U
rH C
cno)
rH «
Q) >i-P
GAG
Q) 4J (0
rH 0) 4->
>i O
X5 i tO >i
O-H O
a) o)
Q)
(ij |J n«
O (C O
Q) Q) Q)
tO 4-> 10
3 U 3
<0 (0 (0
o > o
0) 0)
0
O O M
tj ^^ tj ^}
0) i4J-H
f\ ^j
*M r^
'O '"•^''O O
0) 0) tO
-P -P -P ^»
IT) c m to
ft ft
U r4 U
•H 0) -H i-P
O rH O Q)
(0 Q) ffJ JS
ft-P ft-P
g ffJ g
•H g-H C
•H 0
rH X rH
ff) O (0 73
3 r4 3 ft-P 10
O ft O «J
< < < m
03 43 O 73
•
CO
CO
cn
H
VO
CO
cn
rH
^
0)
tfl
ra
CQ
(0
-p
It)
Q
C
o
•H
4J
n)
N
•H
_J
r^
•H
^J
0)
w
rH
n)
•H
o
0)
€
jjj*
0
o
0)
^^J
•^
X
0
0)
C
0)
rH
J^
^
^J
PQ
• •
0)
o
3
o
CQ
1-3
-------
TABLE 1-2. NATIONWIDE REGULATORY ALTERNATIVE COST IMPACTS
Regulatory
Alternative
A
B
C
D
E
Emission
reduction, %
99
' 97
94
90
89
Total annual
costs, $/MM
12
9.2
6.4
4.3
3.8
Emission reduction,
Mg/yr (tons/yr)
1,061 (1,170)
1,042 (1,149)
1,004(1,107)
963 (1,062)
952 (1,049)
Cost effectiveness
$/Mg ($/ton)
11,300(10,300)
8,800(8,000)
6,400 (5,800)
4,500 (4,050)
4,000 (3,600)
Incremental cost
effectiveness, $/Mg
, ($/ton)
147,000 (133,000)
74,000 (67,000)
51,000(46,000)
45,000 (41,000)
N/A
*Ethylene oxide use cutoff same as for sterilizer vent.
"Status quo means that baseline chambers exhaust emissions are not exceeded.
1-4
-------
production costs will not increase substantially, the price
increases required to recover control also will be low.
In the contract sterilization industry, sterilization is
nearly the entire product, and thus is the main source of
revenue. Therefore, increased sterilization costs due to the
regulation may cause total production costs and prices to
increase significantly. However, as a result of the regulation,
these facilities should experience an increase in demand for
their services through facilities switching from in-house
sterilization to contract sterilization. This increase in demand
should allow them to recover control costs and may even increase
profits for facilities in the contract sterilization industry.
Thus, contract sterilizers are not adversely impacted by the
regulation of commercial sterilizers.
1-5
-------
-------
2.0 INTRODUCTION
2.1 BACKGROUND AND AUTHORITY FOR STANDARDS
According to industry estimates, more than 2.4 billion
pounds of toxic pollutants were emitted to the atmosphere in 1988
("Implementation Strategy for the Clean Air Act Amendments of
1990," Environmental Protection Agency [EPA] Office of Air and
Radiation, January 15, 1991). These emissions may result in a
variety of adverse health effects, including cancer, reproductive
effects, birth defects, and respiratory illnesses. Title III of
the 1990 Amendments to the Clean Air Act provides the tools for
controlling emissions of these pollutants. Emissions from both
large and small facilities that contribute to air toxics problems
in urban and other areas will be regulated. The primary
consideration in establishing national industry standards must be
demonstrated technology. Before national emission standards for
hazardous air pollutants (NESHAP) are proposed as Federal
regulations, air pollution prevention and control methods are
examined in detail with respect to their feasibility,
environmental impacts, and costs. Various control options based
on different technologies and degrees of efficiency are examined,
and a determination is made regarding whether the various control
options apply to each emissions source or if dissimilarities
exist between the sources. In most cases, regulatory
alternatives are subseguently developed and are then studied by
EPA as a prospective basis for a standard. The alternatives are
investigated in terms of their impacts on the environment, the
economics and well-being of the industry, the national economy,
and energy and other impacts. This document summarizes the
information obtained through these studies so that interested
2-1
-------
persons will be able to evaluate the information considered by
EPA in developing the proposed standards.
National emission standards for hazardous air pollutants for
new and existing sources are established under Section 112 of the
Clean Air Act as amended in 1990 [42 U.S.C. 7401 et seq., as
amended by PL 101-549, November 15, 1990], hereafter referred to
as the Act. Section 112 directs the EPA Administrator to
promulgate standards that "require the maximum degree of
reduction in emissions of the hazardous air pollutants subject to
this section (including a prohibition of such emissions, where
achievable) that the Administrator, taking into consideration the
cost of achieving such emission reductions, and any non-air
quality health and environmental impacts and energy requirements,
determines is achievable ... . •' The Act allows the Administrator
to set standards that "distinguish among classes, types, and
sizes of sources-within a category or subcategory."
The Act differentiates between major sources and area
sources. A major source is defined as "any stationary source or
group of stationary sources located within a contiguous area and
under common control that emits or has the potential to emit
considering controls, in the aggregate, 10 tons per year or more
of any hazardous air pollutant or 25 tons per year or more of any
combination of hazardous air pollutants." The Administrator,
however, may establish a lesser quantity cutoff to distinguish
between major and area sources. The level of the cutoff is based
on the potency, persistence, or other characteristics or factors
of the air pollutant. An area source is defined as "any
stationary source of hazardous air pollutants that is riot a major
source." For new sources, the 1990 Amendments state that the
"maximum degree of reduction in emissions that is deemed
achievable for new sources in a category or subcategory shall not
be less stringent than the emission control that is achieved in
practice by the best controlled similar source, as determined by
the Administrator." Emission standards for existing sources "may
be less stringent than the standards for new sources in the same
2-2
-------
category or subcategory but shall not be less stringent, and may
be more stringent than—
(A) the average emission limitation achieved by the best
performing 12 percent of the existing sources (for which the
Administrator has emissions information), excluding those sources
that have, within 18 months before the emission standard is
proposed or within 30 months before such standard is promulgated,
whichever is later, first achieved a level of emission rate or
emission reduction which complies, or would comply if the source
is not subject to such standard, with the lowest achievable
emission rate (as defined by Section 171) applicable to the
source category and prevailing at the time, in the category or
subcategory for categories and subcategories with 30 or more
sources, or
(B) the average emission limitation achieved by the best
performing five sources (for which the Administrator has or could
reasonably obtain emissions information) in the category or
subcategory for categories or subcategories with fewer than
3 0 sources."
The Federal standards are also known as "MACT" standards and
are based on the maximum achievable control technology previously
discussed. The MACT standards may apply to both major and area
sources, although the existing source standards may be less
stringent than the new source standards, within the constraints
presented above. The MACT is considered to be the basis for the
standard, but the Administrator may promulgate more stringent
standards that have several advantages. First, they may help
achieve long-term cost savings by avoiding the need for more
expensive retrofitting to meet possible future residual risk
standards, which may be more stringent (discussed in
Section 2.6). Second, Congress was clearly interested in
providing incentives for improving technology. Finally, in the
1990 Amendments, Congress gave EPA a clear mandate to reduce the
health and environmental risk of air toxics emissions as quickly
as possible.
2-3
-------
For area sources, the Administrator may "elect to promulgate
standards or requirements applicable to sources in such
categories or subcategories which provide for the use of
generally available control technologies or management practices
by such sources to reduce emissions of hazardous air pollutants."
These area source standards are also known as "GACT" (generally
available control technology) standards, although MACT may be
applied at the Administrator's discretion, as discussed
previously.
The standards for hazardous air pollutants (HAP's), like the
new source performance standards (NSPS) for criteria pollutants
required by Section 111 of the Act (42 U.S.C. 7411), differ from
other regulatory programs required by the Act (such as the new
source review program and the prevention of significant
deterioration program) in that NESHAP and NSPS are national in
scope (versus site-specific). Congress intended for the NESHAP
and NSPS programs to provide a degree of uniformity to State
regulations to avoid situations where some States may attract
industries by relaxing standards relative to other States.
States are free under Section 116 of the Act to establish
standards more stringent than Section 111 or 112 standards.
Although NESHAP are normally structured in terms' of
numerical emissions limits, alternative approaches are sometimes
necessary. In some cases, physically measuring emissions from a
source may be impossible or at least impracticable due to
technological and economic limitations. Section 112(h) of the
Act allows the Administrator to promulgate a design, equipment,
work practice, or operational standard, or combination thereof,
in those cases where it is not feasible to prescribe or enforce
an emissions standard. For example, emissions of volatile
organic compounds (many of which may be HAP's, such as benzene)
from storage vessels for volatile organic liquids are greatest
during tank filling. The nature of the emissions (i.e, high
concentrations for short periods during filling and low
concentrations for longer periods during storage) and the
i
configuration of storage tanks make direct emission measurement
2-4
-------
impractical. Therefore, the MACT or GACT standards may be based
on equipment specifications.
Under Section 112(h)(3), the Act also allows the use of
alternative equivalent technological systems: "If, after notice
and opportunity for comment, the owner or operator of any source
establishes to the satisfaction of the Administrator that an
alternative means of emission limitation" will reduce emissions
of any air pollutant at least as much as would be achieved under
the design, equipment, work practice, or operational standard,
the Administrator shall permit the use of the alternative means.
Efforts to achieve early environmental benefits are
encouraged in Title III. For example, source owners and
operators are encouraged to use the Section 112(i)(5) provisions,
which allow a 6-year compliance extension of the MACT standard in
exchange for the implementation of an early emission reduction
program. The owner or operator of an existing source must
demonstrate a 90-percent emission reduction of HAP's (or
95 percent if the HAP's are particulates) and meet an alternative
emission limitation, established by permit, in lieu of the
otherwise applicable MACT standard. This alternative limitation
must reflect the 90- (95-) percent reduction and is in effect for
a period of 6 years from the compliance date for the otherwise
applicable standard. The 90- (95-) percent early emission
reduction must be achieved before the otherwise applicable
standard is first proposed, although the reduction may be
achieved after the standard's proposal (but before January 1,
1994) if the source owner or operator makes an enforceable
commitment before the proposal of the standard to achieve the
reduction. The source must meet several criteria to qualify for
the early reduction standard, and Section 112(i)(5)(A) provides
that the State may require additional reductions.
2.2 SELECTION OF POLLUTANTS AND SOURCE CATEGORIES
As amended in 1990, the Act includes a list of 189 HAP's.
Petitions to add or delete pollutants from this list may be
submitted to EPA. Using this list of pollutants, EPA will
publish a list of source categories (major and area sources) for
2-5
-------
which emission standards will be developed. Within 2 years of
enactment (November 1992), EPA will publish a schedule
establishing dates for promulgating these standards. Petitions
also may be submitted to EPA to remove source categories from the
list. The schedule for standards for source categories will be
determined according to the following criteria:
11 (A) the known or anticipated adverse effects of such
pollutants on public health and the environment;
(B) the quantity and location of emissions or reasonably
anticipated emissions of hazardous air pollutants that each
category or subcategory will emit; and
(C) the efficiency of grouping categories or subcategories
according to the pollutants emitted, or the processes or
technologies used."
After the source category has been chosen, the types of
facilities within the source category to which the standard will
apply must be determined. A source category may have several
facilities that cause air pollution, and emissions from these
facilities may vary in magnitude and control cost. Economic
studies of the source category and applicable control technology
may show that air pollution control is better served by applying
standards to the more severe pollution sources. For this reason,
and because there is no adequately demonstrated system for
controlling emissions from certain facilities, standards often do
not apply to all facilities at a source. For the same reasons,
the standards may not apply to all air pollutants emitted. Thus,
although a source category may be selected to be covered by
standards, the standards may not cover all pollutants or
facilities within that source category.
2.3 PROCEDURE FOR DEVELOPMENT OF NESHAP
Standards for major and area sources must (1) realistically
reflect MACT or GACT; (2) adequately consider the cost, the non-
air quality health and environmental impacts, and the energy
requirements of such control; (3) apply to new and existing
sources; and (4) meet these conditions for all variations of
industry operating conditions anywhere in the country.
2-6
-------
The objective of the NESHAP program is to develop standards
to protect the public health by requiring facilities to control
emissions to the level achievable according to the MACT or GACT
guidelines. The standard-setting process involves three
principal phases of activity: (l) gathering information,
(2) analyzing the information, and (3) developing the standards.
During the information-gathering phase, industries are
questioned through telephone surveys, letters of inquiry, and
plant visits by EPA representatives. Information is also
gathered from other sources, such as a literature search. Based
on the information acquired about the industry, EPA selects
certain plants at which emissions tests are conducted to provide
reliable data that characterize the HAP emissions from well-
controlled existing facilities.
In the second phase of a project, the information about the
industry, the pollutants emitted, and the control options are
used in analytical studies. Hypothetical "model plants" are
defined to provide a common basis for analysis. The model plant
definitions, national pollutant emissions data, and existing
State regulations governing emissions from the source category
are then used to establish "regulatory alternatives." These
regulatory alternatives may be different levels of emissions
control or different degrees of applicability or both.
The EPA conducts studies to determine the cost, economic,
environmental, and energy impacts of each regulatory alternative.
From several alternatives, EPA selects the single most plausible
regulatory alternative as the basis for the NESHAP for the source
category under study.
In the third phase of a project, the selected regulatory
alternative is translated into standards, which, in turn, are
written in the form of a Federal regulation. The Federal
regulation limits emissions to the levels indicated in the
selected regulatory alternative.
As early as is practical in each standard-setting project,
EPA representatives discuss the possibilities of a standard and
the form it might take with members of the National Air Pollution
2-7
-------
Control Techniques Advisory Committee, which is composed of
representatives from industry, environmental groups, and State
and local air pollution control agencies. Other interested
parties also participate in these meetings.
The information acquired in the project is summarized in the
background information document (BID). The BID, the proposed
standards, and a preamble explaining the standards are widely
circulated to the industry being considered for control,
environmental groups, other government agencies, and offices
within EPA. Through this extensive review process, the points of
view of expert reviewers are taken into consideration as changes
are made to the documentation.
A "proposal package" is assembled and sent through the
offices of EPA Assistant Administrators for concurrence before
the proposed standards are officially endorsed by the EPA
Administrator. After being approved by the EPA Administrator,
the preamble and the proposed regulation are published in the
Federal Register.
The public is invited to participate in the standard-setting
process as part of the Federal Register announcement of the
proposed regulation. The EPA invites written comments on the
proposal and also holds a public hearing to discuss the proposed
standards with interested parties. All public comments are
summarized and incorporated into a second volume of the BID. All
information reviewed and generated in studies in support of the
standards is available to the public in a "docket" on file in
Washington, D.C. Comments from the public are evaluated, and the
standards may be altered in response to the comments.
The significant comments and EPA's position on the issues
raised are included in the preamble of a promulgation package,
which also contains the draft of the final regulation. The
regulation is then subjected to another round of internal EPA
review and refinement until it is approved by the EPA
Administrator. After the Administrator signs the regulation, it
is published as a "final rule" in the Federal Register.
2-8
-------
2.4 CONSIDERATION OF COSTS
The requirements and guidelines for the economic analysis of
proposed NESHAP are prescribed by Presidential Executive
Order 12291 (EO 12291) and the Regulatory Flexibility Act (RFA).
The EO 12291 requires preparation of a Regulatory Impact
Analysis (RIA) for all "major" economic impacts. An economic
impact is considered to be major if it satisfies any of the
following criteria:
1. An annual effect on the economy of $100 million or more;
2. A major increase in costs or prices for consumers;
individual industries; Federal, State, or local government
agencies; or geographic regions; or
3. Significant adverse effects on competition, employment,
investment, productivity, innovation, or on the ability of United
States-based enterprises to compete with foreign-based
enterprises *in domestic or export markets.
An RIA describes the potential benefits and costs of the
proposed regulation and explores alternative regulatory and
nonregulatory approaches to achieving the desired objectives. If
the analysis identifies less costly alternatives, the RIA
includes an explanation of the legal reasons why the less costly
alternatives could not be adopted. In addition to requiring an
analysis of the potential costs and benefits, EO 12291 specifies
that EPA, to the extent allowed by the Act and court orders,
demonstrate that the benefits of the proposed standards outweigh
the costs and that the net benefits are maximized.
The RFA requires Federal agencies to give special
consideration to the impact of regulations on small businesses,
small organizations, and small governmental units. If the
proposed regulation is expected to have a significant impact on a
substantial number of small entities, a regulatory flexibility
analysis must be prepared. In preparing this analysis, EPA takes
into consideration such factors as the availability of capital
for small entities, possible closures among small entities, the
increase in production costs due to compliance, and a comparison
2-9
-------
of the relative compliance costs as a percent of sales for small
versus large entities.
The prime objective of the cost analysis is to identify the
incremental economic impacts associated with compliance with the
standards based on each regulatory alternative compared to
baseline. Other environmental regulatory costs may be factored
into the analysis wherever appropriate. Air pollutant emissions
may cause water pollution problems, and captured potential air
pollutants may pose a solid waste disposal problem. The total
environmental impact of an emission source must, therefore, be
analyzed and the costs determined whenever possible.
A thorough study of the profitability and price-setting
mechanisms of the industry is essential to the analysis so that
an accurate estimate of potential adverse economic impacts can be
made for proposed standards. It is also essential to know the
capital requirements for pollution control systems already placed
on plants so that the additional capital requirements
necessitated by these Federal standards can be placed in proper
perspective. Finally, it is necessary to assess the availability
of capital to provide the additional control equipment needed to
meet the standards.
2.5 CONSIDERATION OF ENVIRONMENTAL IMPACTS
Section 102(2)(C) of the National Environmental Policy Act
(NEPA) of 1969 requires Federal agencies to prepare detailed
environmental impact statements on proposals for legislation and
other major Federal actions significantly affecting the quality
of the human environment. The objective of NEPA is to build into
the decision-making process of Federal agencies a careful
consideration of all environmental aspects of proposed actions.
In a number of legal challenges to standards for various
industries, the United States Court of Appeals for the District
of Columbia Circuit has held that environmental impact statements
need not be prepared by EPA for proposed actions under the Act.
Essentially,' the Court of Appeals has determined that the best
system of emissions reduction requires the Administrator to take
into account counterproductive environmental effects of proposed
2-10 '
-------
standards as well as economic costs to the industry. On this
basis, therefore, the Courts established a narrow exemption from
NEPA for EPA determinations.
In addition to these judicial determinations, the Energy
Supply and Environmental Coordination Act (ESECA) of 1974
(PL-93-319) specifically exempted proposed actions under the Act
from NEPA requirements. According to Section 7(c)(l), "No action
taken under the Clean Air Act shall be deemed a major Federal
action significantly affecting the quality of the human
environment within the meaning of the National Environmental
Policy Act Of 1969" (15 U.S.C. 793(c)(l)).
Nevertheless, EPA has concluded that preparing environmental
impact statements could have beneficial effects on certain
regulatory actions. Consequently, although not legally required
to do so by Section 102(2)(C) of NEPA, EPA has adopted a policy
requiring that environmental impact statements be prepared for
various regulatory actions, including NESHAP developed under
Section 112 of the Act. This voluntary preparation of
environmental impact statements, however, in no way legally
subjects the EPA to NEPA requirements.
To implement this policy, a separate section is included in
this document that is devoted solely to an analysis of the
potential environmental impacts associated with the proposed
standards. Both adverse and beneficial impacts in such areas as
air and water pollution, increased solid waste disposal, and
increased energy consumption are discussed.
2.6 RESIDUAL RISK STANDARDS
Section 112 of the Act provides that 8 years after MACT
standards are established (except for those standards established
2 years after enactment, which have 9 years), standards to
protect against the residual health and environmental risks
remaining must be promulgated, if necessary. The standards would
be triggered if more than one source in a category or subcategory
exceeds a maximum individual risk of cancer of 1 in 1 million.
These residual risk regulations would be based on the concept of
providing an "ample margin of safety to protect public health."
2-11
-------
The Administrator may also consider whether a more stringent
standard is necessary to prevent—considering costs, energy,
safety, and other relevant factors—an adverse environmental
effect. In the case of area sources controlled under GACT
standards, the Administrator is not required to conduct a
residual risk review.
2-12
-------
3.0 ETHYLENE OXIDE STERILIZATION/FUMIGATION PROCESSES
AND EMISSIONS
3.1 BACKGROUND INFORMATION
The commercial sterilization (CS) source category covers the
use of ethylene oxide (EO) as a sterilant/fumigant in the
production of medical equipment supplies and in miscellaneous
sterilization and fumigation operations. Commercial
sterilization facilities use EO as a sterilant for heat- or
moisture-sensitive materials or as a fumigant to control
microorganisms or insects. A variety of materials are sterilized
or fumigated with EO, including medical equipment (e.g., syringes
and surgical gloves), spices, cosmetics, and Pharmaceuticals.
These materials may be sterilized at the facility that produces
or uses the product or by contract sterilizers (i.e., firms under
contract to sterilize products manufactured by other companies).
Libraries and museums use EO to fumigate books and other
historical items. State departments of agriculture control
diseases of bees by fumigating beehives with EO.
Information about facilities that use EO as a
sterilant/fumigant was obtained from two sources: (1) a survey
of medical equipment suppliers (Health Industry Manufacturers'
Association [HIMA] members) conducted by HIMA in November 1985
and (2) an information collection request (ICR) submitted by EPA
under Section 114 of the Act to miscellaneous sterilizers and
fumigators (identified during an extensive survey of potential
users) in July 1986. A total of 203 CS facilities responded to
the HIMA survey and the July 1986 EPA information request to
complete the 1986 data base.1'2
3-1
-------
Additional information to the 1986 data base was obtained
from two Section 114 letters (July 1988 and July 1989). The
July 1988 Section 114 letter was sent to 44 (9 parent companies)
of the 203 facilities represented in the 1986 data base.3 These
44 facilities were chosen because they represent the diversity of
sterilizer chamber sizes, annual EO use, and industries
associated with the commercial sterilization category. Although
these facilities represent only 22 percent of the number of
facilities in the CS data base, the emissions from these
facilities account for 64 percent of the total emissions from
commercial sterilization facilities. The July 1988 Section 114
letter was used to obtain detailed operating parameters for a
short-term health risk assessment analysis; data on vacuum pumps,
gas types, control devices, and aeration rooms were also obtained
from this ICR. The July 1989 Section 114 letter was sent to 39
of the 203 facilities in the 1986 data base (i.e., those with a
maximum individual risk [MIR] of cancer incidence greater than
10~3).4 The purpose of this Section 114 letter was to update
EPA's Air Toxics Exposure and Risk Information System (ATERIS)
data base. The July 1989 Section 114 letter was also used to
obtain updated information regarding EO use, emission controls,
and vacuum pumps. The responses to the July 1989 Section 114
letter indicated that 7 of the 39 facilities had ceased EO use.
Therefore, 196 facilities comprise the EPA 1989 CS data base.
As shown in Table 3-1, the facilities represented in the EPA
commercial sterilization data base are located in 41 States and
Puerto Rico. These facilities were grouped by Standard
Industrial Classification (SIC) into the following categories:
1. medical equipment suppliers;
Pharmaceuticals;
other health-related industries;
spice manufacturers;
contract sterilizers;
libraries, museums, and archives;
2
3
4,
5,
6,
3-2
-------
TABLE 3-1.
LOCATIONS OF FACILITIES—EPA COMMERCIAL
STERILIZATION DATA BASE1'4
State
No. of
facilities
State
No. of
facilities
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Florida
Georgia
Illinois
Indiana
Iowa
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Subtotal
3 Missouri
2 New Hampshire
19 New Jersey
3 New York
6 North Carolina
2
5 Ohio
4 Pennsylvania
8 Puerto Rico
4 Rhode Island
3 South Carolina
5 Tennessee
9 Texas
8 Utah
6 Virginia
2 Washington
89 Subtotal
5
2
17
13
7
2
9
14
2
2
3
12
1
5
96
The commercial sterilization data base includes one facility
located in each of the following States: Alabama, Hawaii,
Kentucky, Maine, Nevada, New Mexico, North Dakota, Oregon, South
Dakota, Wisconsin, and West Virginia.
Subtotal
Total No. of
facilities
11
196
3-3
-------
7. laboratories (research, testing, and animal breeding);
and
8. State departments of agriculture.5
Table 3-2 shows the number of facilities in the EPA commercial
sterilization data base for the eight categories listed above and
the SIC codes represented by these industry categories.
3.2 PROCESS DESCRIPTION
There are two main types of EO sterilization processes:
(1) bulk sterilization and (2) single-item sterilization. These
processes are described below followed by discussions about spice
fumigators, library and museum fumigators, and beehive fumigators
used by State departments of agriculture. (In addition to using
the bulk sterilization process, one facility also sterilizes
within 55-gallon drums.2 This process, which is neither bulk nor
single-item sterilization, is not discussed.)
3.2.1 Bulk Sterilization
Bulk sterilization is the more commonly used EO
sterilization process; 98 percent of the commercial sterilization
facilities represented in the EPA data base use this process.1"4
The products to be sterilized are placed in a sterilization
chamber and are exposed to a sterilant gas at a predetermined
temperature, humidity level, and pressure. The equipment,
sterilant gases, and sterilization cycle used for bulk
sterilization processes are described below.
3.2.1.1 Equipment. A schematic of a gas sterilizer is
shown in Figure 3-1. The main components of the sterilizer are
the chamber and vacuum pump. Chambers used by commercial
sterilization facilities typically range in volume from 2.8 cubic
meters (m3) (100 cubic feet [ft3]) to 28 m3 (1,000 ft3);.1'2
Table 3-3 presents size data for the chambers in the EPA
commercial sterilization data base.
A vacuum pump is used to remove air from the chamber before
sterilization begins and to evacuate the sterilant gas after the
sterilization cycle is complete. In the past, once-through,
water-ring vacuum pumps were used. However, many facilities are
converting to full sealant recovery (i.e., oil-sealed or
3-4
-------
TABIjE 3-2. NUMBER OF FACILITIES AND STANDARD INDUSTRIAL
CLASSIFICATION (SIC) PER INDUSTRY CATEGORY—
EPA COMMERCIAL STERILIZATION DATA BASE1"
Industry category
Medical equipment suppliers
Pharmaceuticals
Other health-related
industries
Spice manufacturers
Contract sterilizers
Libraries, museums, and
archives
Laboratories (research,
testing and animal breeding)
State departments of
agriculture
Total
No. of
facilities
61
39
24
23
17
13
11
8
196
SIC
3841, 3842
2834, 5122, 2831,
2833
3079, 3693, 5086,
2211, 2821, 2879,
3069, 3569, 3677,
3999
2099, 5149, 2034,
2035, 2046
7399, 7218, 8091
8411, 8231
0279, 7391, 8071,
8922, 7397
9641
3-5
-------
W
0) -H
N Q
•H
rH Q)
•H -O
M C
a) -H
•p »J
u
(0 C
(0 O
(0 m
o o
•H U
4->
tO 0)
B -a
Q) -H
•s-s
CO (0
U
(0
0)
3-6
-------
TABLE 3-3. CHAMBER SIZES—EPA COMMERCIAL
STERILIZATION DATA BASE1'2
Size range,
m3 (ft3)
<1.4(<50)
1.5-2.8
(51-100)
2.9-14
(101-500)
15-28
(501-1,000)
29-57
(1,001-2,000)
>58 (>2,001)
No. of
chambers
83
29
113
116
55
11
Percent
20
7
28
29
14
2
Cumulative
No. of
chambers
83
112
225
341
396
407a
Cumulative
percent
20
28
55
84
98
100
aThis number excludes four single-item sterilization units,
one 55-gal drum user, and two facilities that did not report
a chamber size.
3-7
-------
recirculating water) vacuum pumps in order to meet the l part per
million by volume (1 ppmv) Occupational Safety and Health
Administration (OSHA) worker exposure standard for EO and
proposed State regulations.3'6
3.2.1.2 Sterilant Gases. Ethylene oxide is an extremely
effective sterilant gas. The EO penetrates product packaging
(e.g., cardboard shipping box, plastic shrink wrap, paper box,
and product wrapping) and destroys bacteria and viruses on the
product. The product remains sterile until use because bacteria
and viruses cannot penetrate the product wrapping.
The most widely used sterilant gas is a mixture of
12 percent by weight EO and 88 percent by weight
dichlorodifluoromethane (CFC-12), referred to as 12/88. Two
other commonly used sterilant gases are (1) pure EO (i.e.,
100 percent EO) and (2) a mixture of 10 percent by weight EO and
90 percent by weight carbon dioxide (CO2), referred to as 10/90.
Other sterilant gas mixtures that are used include 20/8;0, 30/70,
and 80/20 (weight percents EO/CO2).1/2'4 Gas mixtures that
contain 20 percent or greater EO (by weight) are considered
flammable. The 80/20 (EO/CO2) mixture has the same flammability
range as pure EO.7 Physical and chemical properties of EO,
CFC-12, and CO2 are given in Table 3-4. Table 3-5 shows the
number of commercial sterilization facilities represented in the
EPA data base that use a particular gas type and the amount of EO
used for each gas type. Since many commercial sterilization
facilities operate more than one sterilization chamber,! the gas
usage rates in Table 3-5 also are presented on a chamber
basis.1'2'4 The data presented in Table 3-5 are based on
information from the 1985 HIMA survey and the 1986 EPA ICR.1'2
Gas type data were collected for the 44 facilities that; responded
to the 1988 Section 114 letter; however, this information was not
updated for the remaining facilities represented in the 1989 CS
data base. Since 1986, the increased price of CFC-12 (due to EPA
regulation of CFC production) could have affected the number of
facilities that use 12/88.
3-8
-------
w
Q
8
"j
£w
"2
S*
S o
8§
Is
£e
il
«• O
HS
w 9
S§
ss
o
en
Ethylene oxide
£
•a
I
O
"S
1
•a
a
i
*^
«n
-
I
I
u
o
OO
fc
o
vq
U
U
o
•*
O
.
03
•«
.5
a
2
CO
i
s
«
2
.s
a
3-9
-------
l
H
H
to
CQ
r"*t
EH
j(|J
Q
55
O
H
EH
H
H
Cf-t
Cx3
g
tJ
H
CJ
W
§
o
u
H
W
O
g
W
04
W
o
EH
§
H
S
1
in
i
CO
TABLE
**" a
"S O
o ^
^1
* t\
§ s.
o "^
w *£
*^*
5?
1 1
a 1
4) «g
cu w
ti- i2
0|
6 E
"o SS
«-* *^3
g -"I
Cxi **"
'o^k
° ^5
<2
oo
1
•c
CO
«n
o
o\
VO
^cj»
VO
So
cs
f-
V)
c$
1
o
i.
g
?5
i/^
t»-
8.
~
*&
C*i
cn
8
§
o
£
<£
o
o*
V
^f
rl-
O\
C"
2
1
O
o\
0
^4
o\
VO
S
00
VO
1
1
!
6
00
00
O -O
3-10
-------
The 12/88 mixture is popular for several reasons. Unlike
pure EO, 12/88 is nonflammable and nonexplosive. Therefore, the
use of 12/88 does not require explosion-proof rooms and
additional safety precautions that are necessary when pure EO is
used. The 10/90 mixture also is nonflammable and nonexplosive.7
But, because 10/90 is only 10 percent EO by volume whereas 12/88
is 27.3 percent EO by volume, 10/90 requires higher operating
pressures to obtain an EO concentration that is sufficient for
effective sterilization (approximately 304 kilopascals [kPa], or
44 pounds per square inch absolute [psia], for 10/90, as compared
to 170 kPa [24.7 psia] for 12/88).12 The chambers used for
10/90 sterilization must be ASME-rated pressure vessels, (i.e.,
manufactured in accordance with Section VIII, Division I, of the
American Society of Mechanical Engineers [ASME] Pressure Vessel
Code) and are, therefore, more expensive to construct than the
chambers used with 12/88. However, because of insurance
requirements, many commercial sterilization facilities use
chambers that meet requirements for ASME-rated pressure vessels
when sterilizing with 12/88 or with explosive mixtures below
ambient pressure.13
3.2.1.3 Sterilization Cvcle. The typical sterilization
cycle consists of five phases: (1) presterilization
conditioning, (2) sterilization, (3) evacuation, (4) air wash,
(5) chamber exhaust, and (6) aeration. Figures 3-2 and 3-3 show
pressure/time curves for the first four phases of the 12/88
sterilization cycle and the pure EO sterilization cycle,
respectively. Steps 1 through 4 typically require about 8 hours.
3.2.1.3.1 Presterilization conditioning. After the
products have been loaded into the chamber and the airtight door
sealed, a partial vacuum is drawn inside the chamber. This
initial vacuum, or drawdown, prevents dilution of the sterilant
gas. Also, if flammable sterilant gases are used, the removal of
air reduces the potential for ignition.7 The chamber pressure is
reduced to a pressure of about 6.9 to 69 kPa (1 to 10 psia) for
12/88 and 3 kPa (0.4 psia) for pure EO. The initial drawdown
takes from about 5 to 45 minutes, depending on the product being
3-11
-------
UJ
CD
I—I
o
p
I—I
1
•3£
O
I—I
^2
he 1-4
UJ _l
Uo2
UJ UJ
ce. i—
Q. CO
' • •
.-I CM
O
-rH
w
«-H
CO >
CO -H
a)
73
C
0) -
•H C
O O
>1-H
O 4J
m
C b
o o
•H CU
•p M
(0 O
N U
•H
rH 0)
-H -O
^4-H
0) A
-P fc
CO (0
o
• C
CM O
I -H
n C
3 O
ly
•H >i
fe (0
Q)
§
U
3-12
-------
ii
co •
CD
o
o
i eg
§
-H
• (0
O-H
W >
•H
0) Q
H
9 0)
O4-O
Q)
•H
O
O (C
fc
C O
o a
•H M
4J O
(8 U
N
•H Q)
H T3
-H -H
M JQ
0) )-l
4J (0
CO U
• O
n -H
t C
Q)
Q)
.p
^
§
U
3-13
-------
sterilized. Certain products require a longer drawdown time
because they are damaged by sudden pressure changes. The chamber
temperature is then adjusted to between 38° and 54°C (ioo° and
130°F). A higher temperature increases the diffusion rate of EO
into the products and, thus, reduces the time the products must
be exposed to the sterilant gas to ensure proper sterilization.
Finally, the relative humidity is raised to about 45 percent by
injecting steam. Proper humidification is important to| the
process because the susceptibility of microorganisms to the
sterilant gas is increased under moist conditions.7 '
3.2.1.3.2 Sterilization. The sterilant, which is; supplied
as a liquid, is vaporized and introduced into the chamber to
achieve the desired concentration of EO (600 parts per million
[ppm]).7 The chamber pressure depends on the type of sterilant
gas used. Pure EO is used under vacuum pressures of ab^ut 51 kPa
(7.35 psia) ; the 12/88 mixture is used at pressures of iabout
170 kPa (24.7 psia). The pressure is held for about 4 jto
6 hours. This exposure time is dependent on the temperature,
pressure, humidity level, type of sterilant gas, and products
being sterilized. For example, porous products require shorter
exposures than nonporous products. Also, some bacteria are more
resistant to EO and take longer to destroy.
3.2.1.3.3 Evacuation. Following sufficient exposure time,
the sterilant gas is evacuated from the chamber with a vacuum
pump. (If sterilization is performed at a pressure greater than
atmospheric, the chamber is often allowed to vent to atmospheric
pressure before using the vacuum pump to evacuate the chamber.)
Typical evacuation pressures are 13 kPa (1.9 psia) for 12/88 gas
and 3 kPa (0.4 psia) for pure EO. This postcycle vacuum phase
typically lasts about 10 minutes.
3.2.1.3.4 Air wash. The pressure in the chamber is brought
to atmospheric pressure by introducing air (when nonflammable
sterilant gases are used) or either nitrogen or CO2 (when
flammable sterilant gases are used). The combination of
evacuation and air wash phases is repeated from two to four times
3-14
-------
to remove as much of the EO from the product as possible. Each
air wash typically lasts 2 to 15 minutes.
The purpose of the air wash is to allow residual EO to
diffuse from the product. Removal of EO from the product during
the air wash helps meet Food and Drug Administration (FDA)
guidelines on residual EO levels for medical devices, EPA
residual tolerances for agricultural products, and the OSHA
standard for exposure in the workplace.
3.2.1.3.5 Chamber exhaust. Prior to unloading the
steriliser, the chamber door is automatically cracked, and the
chamber exhaust is activated. The chamber exhaust is an exhaust
system that evacuates EO-laden air from the chamber prior to
unloading and while the chamber is being unloaded (and reloaded).
The chamber exhaust typically consists of a butterfly valve in
the ductwork that opens automatically and a roof-mounted blower
that automatically switches on and pulls fresh air through the
chamber. A chamber face velocity of 30.5 m/min (100 ft/min) is
generally maintained, producing a chamber exhaust flow rate of 28
to 85 m3/min (1,000 to 3,000 ft3/min), depending on chamber size.
This process usually begins 15 minutes prior to unloading and
continues during loading and reloading.14
The chamber exhaust is responsible for removing EO from the
void space in the sterilizer chamber, not the product. Use of
the chamber exhaust assists some facilities in meeting the EO
worker exposure levels set by OSHA. Facilities that use
conveyors to load and unload the chamber, as well as facilities
that do not have problems meeting OSHA worker exposure levels,
may not use chamber exhausts.
3.2.1.3.6 Aeration. After the last air wash, the sterile
products are placed in an aeration room and kept there for
several hours to days depending on the product. The purpose of
aeration is to allow further diffusion of residual EO from the
products prior to shipping in order to comply with the FDA and
EPA residual EO guidelines. Ethylene oxide concentrations in the
aeration room are maintained at relatively low levels by
ventilating the room at a rate of about 20 air changes per hour.
3-15
-------
Recent information from industry contacts indicates that
some commercial sterilization facilities are aerating some or all
of the sterile products in heated enclosed aeration units. In
comparison to traditional warehouse-type aeration rooms, these
units are smaller in volume (<70 m3 [2500 ft3]) with much lower
ventilation rates. Consequently, the EO concentrations are
usually higher than the 1 ppmv OSHA standard. However, worker
exposure is reduced by not opening the door until the EO
concentration drops and by limiting the frequency of opening the
aeration room door. The main purpose of this type of aeration
process is to increase the diffusion rate of EO out of the
sterile product (by increasing the temperature) and, thus, reduce
the aeration time. !
Facilities that sterilize products infrequently may aerate
in the sterilization chamber. Two basic chamber aeration
processes are used. The first process involves cycling the
chamber between atmospheric pressure and a slight vacuum pressure
(i.e., a pressure of about 94 kPa [13.7 psia]) several times over
a 12- to 24-hour period. The length of these cycles depends on
the chamber size and vacuum pump capacity. The second process
involves drawing an extreme vacuum (about 0.6 kPa [0.1 psia]) in
the chamber and holding the vacuum for 24 to 48 hours.
Some commercial sterilization facilities with small
sterilizers (less than 1m3 [40 ft3]) use aeration chambers (or
cabinets), which are similar to the sterilization chambers in
size and design. These facilities typically aerate products for
about 24 hours.
3.2.2 Single-Item Sterilization System
Four of the 196 commercial sterilization facilities
represented in the 1989 EPA CS data base (2 percent) reported the
use of a single-item sterilization system.1'2 Three of these
facilities use the Sterijet® system manufactured by
H. W. Andersen Products; one facility uses another patented
system that is similar to the Sterijet® system.2 In contrast to
the bulk sterilization chambers used by most commercial
sterilization facilities, these systems are designed to sterilize
3-16
-------
small individual items (such as medical equipment supplies) in
sealed pouches. Marketing of these systems is primarily focused
on hospital sterilization.15
These single-item sterilization systems consist of (1) a
machine that delivers the sterilant gas through a nozzle,
(2) flexible plastic pouches, and (3) an aeration cabinet. The
process involves the following steps. The product to be
sterilized is placed in a plastic pouch. With the open ends of
the pouch sealed around the nozzle, a slight vacuum is drawn in
the pouch followed by injection of a premeasured quantity of
sterilant gas. The amount of sterilant gas injected depends on
the size of the pouch. After the gas is injected, the nozzle is
automatically withdrawn, and the pouch is heat sealed. The
sealed pouches are placed directly into an aeration cabinet or
temperature-controlled aeration room. The enclosed product is
sterilized prior to the escape of the gas through the pouch,
which is designed to retain the EO long enough to ensure proper
sterilization. The products are sterilized for approximately
12 hours at about 50°C (122°F) and aerated for 36 hours.15
Another type of single item sterilization system consists of
(1) ampule-delivered EO sterilant gas, (2) flexible plastic
pouches, (3) a sterilization cabinet, and (4) an optional
aeration cabinet.16 This process involves the following steps.
The product to be sterilized is wrapped in gas permeable
packaging and placed in a plastic pouch along with an ampule
containing the sterilant gas. The pouch is placed in the
sterilization cabinet, and the ampule is broken to release the
sterilant gas. The pouch is then sealed, and the cabinet closed
to allow sufficient time for sterilization of the materials.
After the sterilization cycle is- complete, the cabinet door and
plastic pouch are opened, and the materials are unloaded.
Depending on the characteristics of the materials sterilized,
they may be placed in an aeration cabinet to allow for the
offgassing of residual EO. The products are typically sterilized
for approximately 12 hours at room temperature and atmospheric
3-17
-------
pressure, and aerated for approximately 12 hours at 50°C (122°F)
at atmospheric pressure (if an aeration cabinet is used).16
At least one vendor does offer an optional or retrofit
ventilation hood to control worker exposure to EO when
loading/unloading the sterilizer, and while sterilization occurs.
This device causes fresh air to be pulled past the chamber by a
roof-mounted blower; however, this system does not provide any
evacuation of the sterilization cabinet itself.16
3.2.3 Spice Fumiaators
The process for spice fumigators is essentially the same as
bulk sterilization.17 The spices are typically stored in fiber
drums lined with a plastic insert, which is closed with twist
tie; lids are then placed on the drums. Alternatively, the
spices may be stored in large bags or totes. The drums, bags, or
totes are loaded into the sterilization chamber on wooden
pallets, typically via a conveyor. Depending on how densely
packed the spice is, a long, hollow spike punctured with many
holes may be driven into the spice to allow the EO to penetrate
through the bag drum, or tote. Ethylene oxide is then added to
the chamber. The length of the sterilization cycle depends on
the product's susceptibility to adequate kill rates. Ethylene
oxide's effectiveness is different for different spices.
Following evacuation of the sterilization chamber, and subsequent
air washes, the spices are removed from the sterilization chamber
and placed in an aeration room. Aeration typically takes
2 hours.17
3.2.4 Library and Museum Fumiaators
Library and museum fumigation is accomplished using
essentially the same process as for bulk sterilization. However,
the amount of ethylene oxide used each year by these facilities
is much lower than that used by the typical bulk sterilizers.
These library and museum fumigators are typically operated only
one to two times a month.18'21 Additionally, several museums and
libraries are discontinuing their use of ethylene oxide for
fumigation because of (1) the recent OSHA worker exposure
regulations and (2) problems associated with the long aeration
3-18
-------
times needed when fumigating organic materials with EO (most.
museums and libraries do not have the space needed for adequate
aeration of organic materials, such as wood [up to l month],
because these materials readily absorb much of the EO used in the
fumigation process).22"24 Some museums have switched to using
sulfuryl fluoride as a fumigarit.22'23 Other museums are
accepting objects from other museums and libraries to be
fumigated with EO in their fumigator.25
3.2.5 Beehive Fumiqators
The process for beehive fumigators is essentially the same
as bulk sterilization; however, a unique feature of the
fumigators warrants a separate discussion. Whereas the
sterilization processes described above are performed at one
location, the eight State departments of agriculture represented
in the EPA sterilization data base use portable chambers to
fumigate beehives at numerous and variable locations in each of
the six States.2 The State departments of agriculture use an
EO/CO2 sterilant gas mixture. Typically, a garden hose is
connected to the fumigation chamber and is placed along the
ground for venting the sterilant gas during the evacuation phase
of the sterilization cycle. After the evacuation, the beehive is
removed from the chamber and aerated in the open air.
3.3 EMISSION SOURCES
The four principal sources of EO emissions from
sterilization/fumigation processes are (1) the sterilizer vent(s)
(i.e., the vent on the vacuum pump gas/liquid separator), (2) the
sterilization chamber vacuum pump drain (assuming that a once-
through, water-ring vacuum pump is used), (3) the chamber exhaust
vent(s), and (4) the aeration room vent(s). A schematic of these
emission sources is shown in Figure 3-4. Other potential
emission sources are equipment leaks and storage and handling.
(For the purposes of developing emission estimates, and because
bulk sterilization processes are the main source of emissions,
emission sources were assumed to be the same for all stationary
sterilization processes [i.e., bulk and single-item and the
55-gallon drums]).
3-19
-------
ETHYLENE.
OXIDE
PRODUCTS FOR
STERILIZATION
(760 Mg/yr) <'.000 Mg/yr) (110 Hg/yr)
VENT
VENT OR EMISSION
CONTROL DEVICE
(GAS)
SEPARATOR
(LIQUID)
DRAIN
STERILIZATION
CHAMBER
STERILIZED,
PRODUCTS
VENT
AERATION
ROOM
Figure 3-4,
Schematic of emission sources at commercial
sterilization facilities.
3-20
-------
3.3.1 Sterilization Chamber Vents
Sterilization chamber vent emissions are dependent on the
type of vacuum pump used to evacuate the sterilizer. Once-
through, liquid-ring design vacuum pumps that use water as the
working fluid discharge a mixture of chamber gas and water to a
centrifugal gas/liquid separator. In the separator, gaseous EO
is directed to a vent and emitted to the atmosphere. The liquids
from the separator are discharged to a drain. (Full-sealant
recovery [i.e., oil-sealed or recirculating water] vacuum pumps
do not produce drain emissions of EO.) Sterilizer vent emissions
also include emissions associated with venting the chamber from a
positive pressure before evacuating with a vacuum pump.
3.3.2 Sterilization Chamber Vacuum Pump Drains
If a once-through, water-ring vacuum pump is used to
evacuate the chamber, some of the EO evacuated from the chamber
enters the liquid-phase separator line with the vacuum pump
water. Although some EO may be hydrolyzed to ethylene glycol,
the conversion rate at ambient temperatures is extremely slow,
requiring weeks for completion (see Figure 3-5).8 Also, EO is
rapidly released from an aqueous solution when agitated.26
Therefore, virtually all of the EO that dissolves in the vacuum
pump water is emitted from the water. The absorbed EO may be
released at the 1-inch air break between the liquid pipe and
drain (required by local plumbing codes) or may diffuse into
other areas of the building as the water passes through the drain
system. Any remaining EO would desorb into the head space of the
sewer pipes (possibly creating flammable mixtures with air) and
be emitted as it passes through the sewer or waste treatment
systems.6'8
3.,3.3 Chamber Exhaust Vent
Chamber exhaust emissions consist of EO that remains in the
sterilizer chamber void space (surrounding the product) after the
sterilization cycle is completed. Product off-gassing in the
sterilizer is a negligible contributor to this emission source.27
Therefore, the chamber exhaust emissions are assumed to be only
EO trapped in the sterilizer void space.
3-21
-------
20
30 40
Time, days
50
60
70
Figure 3-5. Hydrolysis rates of dilute, neutral aqueous
solutions of ethylene oxide. (Courtesy of Union Carbide
Corporation, Ethylene Oxide/Glycol Division.)8
3-22
-------
3.3.4 Aeration Room Vent
All emissions from residual EO in the product are referred
to as aeration room vent emissions. As residual EO diffuses out
of the sterile products in the aeration room or is emitted to the
sterilization room when the chamber door is opened, it is emitted
to the atmosphere via room ventilation. High ventilation airflow
rates are used to maintain EO concentrations below the OSHA limit.
3.3.5 Equipment Leaks
Although equipment component counts (number of flanges,
valves, etc.) were not obtained for the commercial sterilization
facilities, observations made during site visits indicated that
the number of components is small. However, control of equipment
leaks may be important to meet the 1 ppmv OSHA standard. For the
purposes of this analysis, equipment leak emissions are assumed
to be negligible.
3.3.6 Storage and Handling
Ethylene oxide at commercial sterilization facilities is
typically stored in pressurized cylinders rather than bulk
containers. Therefore, material losses associated with loading
and unloading bulk quantities of EO and storage tank breathing
losses would not occur. Although bulk storage of sterilant gas
at sterilization facilities is rare, at least one commercial
sterilization facility stores bulk quantities of 12/88 in a
pressure vessel. During transfer of the 12/88 from the tank
truck to the pressure vessel, the vessel and the tank truck are
vapor balanced. Therefore, emissions during transfer are
expected to be negligible. Also, because the storage tank is a
pressure vessel, no emissions should occur during routine
operation. Consequently, commercial sterilization facilities are
likely to have negligible storage and handling emissions.
3.4 EMISSION ESTIMATES
3.4.1 Commercial Sterilization Facilities
The emission estimate for commercial sterilization
facilities is based on the facility-specific annual EO usages and
emission control levels reported in the 196 responses to the HIMA
survey, the 1986 EPA ICR, and the July 1989 Section 114
3-23
-------
letter.1'2 Eight of the 196 facilities are State departments of
agriculture that operate a total of 10 mobile beehive fumigator
units; these units are not included in these or subsequent
analyses unless otherwise stated. (These 10 fumigation units use
a total of 0.46 megagrams per year (Mg/yr) (1,015 pounds per year
[Ib/yr]) of EO, all of which is reportedly uncontrolled.)
Average EO emissions from the remaining 188 commercial
sterilization facilities, based on total sterilizer volume, are
presented in Table 3-6. The total amount of EO used by these
188 commercial sterilization facilities is 1,920 Mg/yr
(4.23 million Ib/yr); approximately 42 percent (i.e., 809 Mg/yr
I
[1.78 million Ib/yr]) of this amount is controlled. Therefore,
the EO emission estimate for the 188 facilities represented in
the 1989 EPA CS data base is 1,111 Mg/yr (2.45 million Ib/yr).1"4
Of this amount, it is estimated that 667 Mg/yr (1.47 million
Ib/yr) are emitted from sterilizer vents; 312 Mg/yr
(688,000 Ib/yr) are emitted from sterilization chamber vacuum
pump drains; 38 Mg/yr (84,000 Ib/yr) are emitted from chamber
exhaust vents; and 57 Mg/yr (126,000 Ib/yr) are emittedjfrom
aeration room vents (see Figure 3-4). These estimates were
developed using the HIMA survey, the EPA ICR responses, the
Section 114 letter (July 1988 and July 1989) responses, and the
following assumptions:
1. All of the EO used in the sterilization process is
evacuated from the sterilization chamber or released from the
product during postevacuation processes.
2. Within each facility, EO emissions are distributed among
four emission points. The four emission points and the
percentage of total EO emissions allocated to each are:
a. Sterilizer vent(s)—50 percent;
b. Sterilization chamber vacuum pump drain (if
applicable)—45 percent;
c. Chamber exhaust vent(s)—2 percent; and
d. Aeration room vent(s)—3 percent.
This 50/45/2/3-percent split is based on industry estimates,
limited test data, and engineering judgment.28'29
3-24
-------
TABLE 3-6. AVERAGE EMISSIONS FROM COMMERCIAL
STERILIZATION FACILITIES—EPA COMMERCIAL STERILIZATION
DATA BASE1"4
Total chamber volume at
facility, m3 (ft3)
<11 (<400)
11-56 (400-2,000)
>56 (> 2,000)
No. of facilities
87
71
38
Mean EO use, kg/yr (Ib/yr)
580(1,300)
6,500 (14,000)
37,000 (82,000)
Mean EO emissions, kg/yr
Ob/yr)1111
520 (1,200)
4,200 (9,300)
20,000 (45,000)
&Mean emissions are less than mean EO use because of existing controls.
''Emissions from all sources (i.e., sterilizer vent, vacuum pump drain, aeration).
3-25
-------
3. For the uncontrolled sterilization chambers, all of the
EO that enters the chamber (except 3 percent, if the aeration
room is controlled) is released to the atmosphere. For the
79 sterilization chambers with emission control devices, the
chamber vent emissions are assumed to be controlled at the
following efficiencies:
a. 99.0 percent for acid/water scrubbers and catalytic
oxidizers;
I
b. 99.0 percent for flares (given the chemical/ physical
characteristics of ethylene oxide, its high combustabiiity, and
the extremely weak nature of the oxide bond, it is reasonable to
assume that emissions of ethylene oxide will be controlled at an
efficiency of 99 percent in flares rather than the generally EPA-
accepted efficiency for flares of 98 percent); and I
c. The .facility-reported efficiency for other control
devices.
4. All facilities that control sterilizer vent emissions
with acid/water scrubbing, catalytic oxidation, flaring, or
condensation/ reclamation control devices are assumed t'o have
recirculating-fluid vacuum pumps and, thus, no drain emissions
from those chambers. (One facility that uses a different control
technology than those described above is assumed to haye a once-
through water-sealed pump). All uncontrolled facilities are
assumed to have once-through water-sealed pumps unless data are
available to indicate otherwise.
5. At facilities that have once-through water-sealed vacuum
pumps, all of the EO that dissolves in the vacuum pump water
subsequently enters the drain and is assumed to be emitted
uncontrolled to the atmosphere at an outdoor ground-lev!el drain
near the facility. 'This assumption is consistent with test data
that suggest EO is rapidly released from an aqueous solution when
agitated.26
6. Facilities that have a total sterilizer volume! greater
O*5
than 7 m° (>250 ftj) are assumed to have chamber exhausts on all
sterilizers. Chamber exhaust emissions are assumed to equal
3-26
-------
2 percent of the total EO use and to be released uncontrolled to
the atmosphere.29
7. At each facility (except for three facilities that have
aeration room controls), all of the EO that enters the aeration
room(s) vent is released uncontrolled to the atmosphere.
3.5 CURRENT REGULATIONS
3.5.1 Occupational Safety and Health Administration Standard
In 1984, OSHA established a permissible exposure limit for
occupational exposure to EO of 1 ppmv determined as an 8-hour
time-weighted average (TWA) concentration. In addition, an
action level of 0.5 ppmv as an 8-hour TWA was established as the
level above which employers must monitor employee exposure.30 In
April 1988, OSHA established a short-term excursion limit (EL)
for occupational exposure to EO emissions of 5 ppmv averaged over
a 15-minute sampling period.3^-
3.5.2 State Regulations
Existing State regulations for EO are summarized in
Table 3-7. Several States are currently regulating EO or
developing air toxics programs.32"38
3-27
-------
TABLE 3-7. STATE REGULATIONS FOR ETHYLENE OXIDE EMISSIONS22'28
State
California*
Regulatory description
• Control is based on annual EO usage
Annual Sterilizer
usage. Ib control
<25 No control
25-600 99%
600-5,000 99.9%
>5,000 99.9%
•
Aeration
control
No control
No control
95%
99%
Colorado
Connecticut
Florida*
Michigan*
Missouri
New Jersey
New York0
Oklahoma
Puerto Rico
Rhode Islandb
Tennessee
Texas
Utah
Vermont
Virginia
Wyoming
• Regulate as a volatile organic compound (VOC).
• Reasonably available control technology (RACT) required for new sources.
• Best available control technology (BACT) required for all new or modified sources exceeding a
maximum allowable stack concentration (MASC).
• MASC is calculated using exhaust gas flow rate, stack height, and the distance from the discharge
point to the property line. MASC would be exceeded for industrial sterilizers using typical
sterilization cycles. Therefore, BACT required on new or modified sources. Existing sources
exceeding the maximum allowable ambient concentration of 0.01 ppm have 3 years to comply
with orders given by the Connecticut Department of Environmental Protection. !
• Maximum risk level of 10"^ for new or modified sources.
• BACT for all new sources. Requiresuemissions be indectable or subjected to risk analysis
(maximum allowable risk level is 10"°). For industrial sterilizers using typical sterilization cycles,
a control efficiency based on a risk assessment analysis would be greater than 99 percent by
weight. I
• Regulate as a VOC.
• Regulate as a VOC.
• BACT required for new or modified sources.
• New or modified sources must receive 99 percent control or greater, or BACT (also at permit
reviews) ;
• Maximum annual impact must not exceed guideline Acceptable Ambient Level (AAL) of 6.67
/ig/nr (a revised AAL of 0.019 /ig/nr is anticipated for the next edition of Air Guide-1 [to be
released by 01/90].
• Certificate of operation includes the following statement:
"Should significant new scientific evidence from a recognized institution result in the
decision by DEC that lower ambient levels must be established, it may be necessary
to reduce emissions from this source prior to the expiration of this Certification of
Operation."
• Maximum ambient air concentration at property line is 1/100 of TLV.
• Regulate as a VOC.
Emission controls required for emissions greater than 3 Ib/h or 15 Ib/d.
Maximum risk level of 10"* for new and existing sources.
If BACT is used, may consider 10"^ risk level. ]
Regulate under standards for process and nonprocess emissions.
BACT required for all new sources.
BACT required for all new or modified sources. BACT requirements to go into effect for existing
sources. ;
Following the programs developed in New York.
Regulate as a VOC.
For any 24-hour concentration exceeding 1/100 of the TLV-TWA both existing and new facilities
are required to control emissions as directed by the Virginia Air Pollution Control Board.
BACT required for all new sources. ,
Controls must meet AAL at property line.
"Information obtained from State contacts in May 1990. '
"Information obtained from State contacts in February 1989. All other information is from 1986 through 1987
data.
3-28
-------
REFERENCES FOR CHAPTER 3
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Letter and enclosures from Jorkasky, J., Health Industry
Manufacturers Association (HIMA), to D. Markwordt, EPA:CPB.
February 21, 1986. Survey responses from HIMA members.
Memorandum. deOlloqui, V., to Commercial Sterilization
files. August 22, 1990. Responses to the July 1986
Section 114 Information Collection Request.
Memorandum. deOlloqui, V., to Commercial Sterilization
files. August 22, 1990. Responses to the July 1988
Section 114 Letter.
Memorandum. deOlloqui, V., to Commercial Sterilization
files. August 22, 1990. Responses to the July 1989
Section 114 Letter.
Commercial Sterilization Standard Industrial Classification
(SIC) data base. Research Triangle Institute. July 1987.
SIC designations for facilities in the EPA commercial
sterilization data base.
Letter from Buonicore, A., Chemrox, Inc., to D. Markwordt,
EPA:CPB. August 27, 1984. Comments on the sources of
ethylene oxide emissions draft report.
Gas Sterilants. Product information brochure.
Carbide Corp., Linde Division. Undated.
Union
13.
Ethylene Oxide Product Information Bulletin. Union Carbide
Corp., Ethylene Oxide/Glycol Division. 1983.
Ethylene Oxide: Material Safety Data Sheet. General
Electric. April 1983.
Dichlorodifluoromethane: Material Safety Data Sheet.
Genium Publishing Corporation. February 1986.
Handbook of Chemistry and Physics. 67th Edition. CRC
Press, Boca Rotan, FL. 1986.
Telecon. Taylor, G., MRI, with S. Conviser, and C. Waltz,
Union Carbide Corporation, Linde Division. July 31, 1987.
Discussion of operating pressures for sterilization
chambers.
Letter from Burley, R., Environmental Tectonics Corporation,
to S. Wyatt, EPA: CPB. August 25, 1987. Comments on draft
BID Chapter 3 for ethylene oxide NESHAP.
3-29
-------
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
Buonicore, A. J. In-plant Programs to Reduce Ethylene Oxide
Worker Exposure Levels. International Society of
Pharmaceutical Engineers. Volume 4, No. 4. July through
August 1984.
Mitigation of Worker Exposure to Ethylene Oxide.
Goldgraben, R. and N. Zank, The Mitre Corp. 1981.
Anpro. Product Information Brochures.
Products, Inc., Undated.
H. W. Anderson
Trip Report. Beall, C., MRI, to D. Markwordt, EPArCPB.
McCormick and Company, Inc., Hunt Valley, Maryland.
January 16, 1986.
Telecon. Beall, C., MRI, with R. Jones. North Carolina
Archives and Records, Department of Cultural Resources.
November 19, 1985.
Telecon. Friedman, E., MRI, with R. Hauser. Old Dartmouth
Historical Society: Whaling Museum, New Bedford, MA.
February 28, 1986.
Telecon. Friedman, E., MRI, with J. Brown. Lowie; Museum of
Anthropology: University of California. Berkeley, CA.
March 19, 1986.
Telecon. Newton, D., MRI, with S. Walker.
Library. Boston, MA. February 28, 1986.
Boston Public
Telecon. Beall, C., MRI, with V. Wilcox. Smithsonian
Museum: Museum Support Center. Washington, DC. |
November 21, 1985.
Telecon. Beall, C., MRI, with L. Zycherman. Museum of
Modern Art: American Institute of Conservation. -New York,
NY. December 19, 1985.
Telecon. Beall, C., MRI, with W. Jessup. Smithsonian
Museum: Museum Support Center. Washington, D.C.,
November 27, 1985.
Trip Report. Beall, C., MRI, to D. Markwordt, EPA:CPB.
North Carolina Archives and Records: Department of Cultural
Resources. Raleigh, NC. December 6, 1985.
Conway, R., G. Wagg. M. Spiegel and R. Berglund.
Environmental Fate and Effects of Ethylene Oxide.
Environmental Service and Technology. 1983. 17(2):107-112.
Memorandum. deOlloqui, V., and S. Srebro. MRI, to
D. Markwordt, EPA/CPB. March 21, 1990. Costing of Control
Alternatives for the Rear Chamber Exhaust.
3-30
-------
28
29
30
31
32
33
34
35
36
37
38
Abrams, w. , McCormick and Company, Inc. Project No. 075320,
Treatment of Spices-EtO Mass Balance. Final Report.
November 26, 1985.
Memorandum. Srebro, s., MRI, to D. Markwordt., EPA/CPB.
March 21, 1991. Baseline Cost of Reducing Ethylene Oxide
Emissions from Sterilizer Vents and Associated Vacuum Pump
Drains.
Ethylene Oxide. Occupational Safety and Health
Administration. Promulgated on June 22, 1984. 49 FR 25797.
Office of the Federal Register. Washington, DC.
Occupational Exposure to Ethylene Oxide. Occupational
Safety and Health Administration. Promulgated on April 6,
1988. 29 CFR Part 1910. Office of the Federal Register.
Washington, DC.
Summary of Regulations Pertaining to Ethylene Oxide by
State. Chemrox, Inc. Bridgeport, CT. Undated.
Air Pollution Control. The Bureau of National Affairs, Inc.
Washington, DC. January 1987.
Telecon. Shine, B., MRI, with R. Vincent, California Air
Resources Board. February 14, 1989.
Telecon. Shine, B., MRI, with J. Glenn, Florida Department
of Environmental Regulation, Division of Air Resources
Management. February 14, 1989.
Telecon. Shine, B. , MRI, with P. Schleusener, Michigan
Department of Natural Resources, Air Quality Division.
February 14, 1989.
Telecon. Shine, B. , MRI, with E. Wade, New York Department
of Environmental Conservation Division of Air Quality.
February 14, 1989.
Telecon. Shine, B., MRI, with B. Morin, Rhode Island
Department of Environmental Management, Division of Air and
Hazardous Materials. February 14, 19.89.
3-31
-------
-------
4.0 EMISSION CONTROL TECHNIQUES
This chapter describes the techniques available to control
ethylene oxide (EO) emissions from bulk sterilization and single-
item sterilization processes. Alternatives to EO sterilization,
retrofit considerations, and the impacts of chlorofluorocarbon
(CFC) regulation on EO emission controls are also discussed.
4.1 BULK STERILIZATION PROCESSES
Discussed below are techniques available to control
emissions of ethylene oxide (EO) from the four principal sources
of emissions from bulk sterilization processes:
1. The sterilizer vent(s) (i.e., the vent on the vacuum
pump gas/liquid separator);
2. The sterilization chamber vacuum pump drain;
3. The chamber exhaust vent; and
4. The aeration room vent.
4'1'1 Sterilization Chamber Vent Emissions
Three primary techniques are used to control EO emissions
from sterilizer vents: hydrolysis, oxidation, and condensation.
Ethylene oxide is catalytically hydrolyzed to form ethylene
glycol; thermal or catalytic oxidation decomposes EO into carbon
dioxide and water; and condensation allows the recovery of the
sterilant gas mixture. A fourth control technique for sterilizer
vents is a gas/solid reactor system that chemically reacts with
EO and binds it to the solid packing of the reactor.1
Table 4-1 shows the emission control techniques and devices
for sterilizer vent emissions used by the controlled facilities
represented in the EPA data base (refer to Chapter 3 for a
description of the contents and origin of the data base). Forty
of the 188 commercial sterilization facilities (21 percent) in
4-1
-------
CO
CO
J H
« r
S
ts &
oo
o\
o J2
8
s;
«^
g
PQ
H
CO
8-8
Q Q
1%
•A
o\
cs
Ov
00 _
8
8
H
u
tSJ
H
vo
a
o\
8
o\
o\
£ M
S-S
I
S
J5
"8
pi £
u.
•S
o o
11
4-2
-------
the data base reported the use of a .control device for sterilizer
vent emissions. Twenty-eight of these 40 facilities use one
emission control device for multiple chambers by manifolding the
chamber vents and staggering the evacuation of the sterilant gas
from the chambers. The remaining 12 facilities control emissions
from single chambers.2'5
Nine additional commercial sterilization facilities reported
the use of a neutral-water scrubber to control EO vent emissions.
Neutral-water scrubbers reduce EO vent emissions by "washing" a
portion of the EO to the drain (facilities reported 20 to
100 percent of the total EO emissions from the sterilizer chamber
were "controlled" by a neutral-water scrubber).2"5 Some of the
EO that is washed to the drain may be converted by hydrolysis to
ethylene glycol; however, the conversion rate of EO to ethylene
glycol in neutral water at ambient temperatures is extremely
slow, requiring weeks for completion.6 Since EO is rapidly
released from an aqueous solution when agitated, the vast
majority of the EO washed to the drain is expected to off-gas
uncontrolled from the air break in the drain line, sewer lines,
or the waste water treatment system.6"8 Because the use of
neutral-water scrubbers merely changes the EO emission source,
these scrubbers are not discussed here as a control technique.
4.1.1.1 Hydrolysis. Hydrolysis is the most common EO
emission control technique used by commercial sterilization
facilities.2"5 This technique is applicable for both pure EO and
EO/inert gas mixtures such as 12/88 (12 percent by weight EO and
88 percent by weight dichlorodifluoromethane [CFC-12]) and 10/90
(10 percent by weight EO and 90 percent by weight carbon dioxide
[C02]).
Ethylene oxide can be hydrolyzed under relatively mild
conditions to ethylene glycol products (without affecting the
inert gas) as shown in the following reaction:
4-3
-------
C2H40
H20
H-f or OH-
HOCH2CH2OH
HO(CH2CH2)nOH
Ethylene
oxide
Ethylene
glycol
Polyethylene
glycols
Ethylene oxide will hydrolyze in neutral water, but this reaction
is very slow. (The half-life of EO in neutral water at ambient
temperatures is approximately 14 days.)8 The reaction irate is
increased in an acidic or basic solution. The reaction is
approximately two orders of magnitude faster under acidic
conditions than under basic conditions, making acid hydrolysis
the preferred method. Twenty-eight of the 188 commercial
sterilization facilities represented in the EPA data bdse
reported using acid-water scrubbers; one facility reported using
caustic scrubbers to control EO emissions.2"5
4.1.1.1.1 Packed scrubbers. Figure 4-1 is a schematic of a
The system
packed scrubbing system used to control EO emissions.
consists of a countercurrent packed tower, a reaction vessel, and
a holding tank. In the countercurrent tower, the sterilant gas
is contacted with an acidic water solution, generally aqueous
sulfuric acid. Because EO is extremely water soluble, most of
the EO is absorbed into the scrubber liquor. Next, the' liquor is
sent to the reactor vessel, which is a small storage tank
operated at atmospheric pressure, to complete the hydrolysis of
EO. After the reaction is complete, the liquor is sent to the
storage vessel. The liquor in the storage vessel is recirculated
to operate the tower until the concentration of the ethylene
glycol in the liquor reaches a predetermined weight percentage,
past which point the scrubber efficiency declines. Manufacturers
of packed scrubbing systems suggest that the scrubbing liquor is
spent when the solution is 30 to 40 percent by weight elbhylene
glycol.9'10 Possible methods of determining when the liquor
needs replacing include liquid level indicators or specific
gravity detectors in the tank. (Both parameters increase as the
amount of ethylene glycol increases.) Alternatively, the amount
of EO charged to the sterilizer can be used to determine the
liquor changeout point. The spent solution is neutralized and
4-4
-------
(0
>i
W
•H
O
CO
13
0)
*t
U
(0
0)
3
U
^
0)
4J
3
O
U
0)
4-5
-------
then disposed or sold. (See Section 4.1.1.1.3 for a more
detailed discussion of waste disposal.) Generally, sodium
hydroxide is used to neutralize the glycol solution; sodium
carbonate can also be used.
Countercurrent packed scrubbers are used by commercial
sterilization facilities with sterilizers ranging from il.l cubic
meters (m3) (40 cubic feet [ft3]) to 170 m3 (6,000 ft3)!.
Ethylene oxide use at these commercial sterilization facilities
ranges from 0.8 megagrams per year (Mg/yr) (2,000 pounds per year
[lb/yr]) to 84 Mg/yr (180,000 lb/yr).2'3'5
Manufacturers of countercurrent packed scrubbers designed to
control EO emissions from sterilizer vents claim EO removal
efficiencies greater than 99 percent. •*-'y'
1/9,11
For a 12/88
sterilant-gas mixture, the average EO removal efficiency for
three tests was 99.0 percent by weight (individual test results
were 99.0, 98.7, and 99.4 percent)
12
These tests were conducted
using a scrubber that was designed to achieve an EO removal
efficiency of 99 percent. A representative from the manufacturer
of the tested acid-water scrubber stated that the company can
design scrubbers to achieve virtually any EO removal eftficiency
with any type of sterilant gas
13
The results of an
EPA-sponsored test on another acid-water scrubber designed by
this company indicated an EO removal efficiency greater; than
99.9 percent for 12/88
14
For pure EO, the EO removal efficiency
review of the
was greater than 99.98 percent for each of four tests performed
at two facilities.12'15 However, a detailed
available test data indicates that 99.0 percent is the highest EO
removal efficiency that can be achieved on a continuous! basis
based on limited data at various EO concentrations.16 i
4.1.1.1.2 Reaction/detoxification towers. Another acid
hydrolysis scrubbing technique for EO emission control is a
reaction, or detoxification, tower. A schematic of this system
is shown in Figure 4-2. The reaction/detoxification tank holds a
sulfuric acid solution of pH 0.5 to 2.5. As the sterilknt gas
bubbles upward through the acidic liquor, EO is absorbed, and
catalytically hydrolyzed to ethylene glycol. The gas stream then
4-6
-------
(Q
JJ
O
O
•I
O
O
•H
4J
(0
O
X
O
4J
0)
Q
0)
^
3
CT
4-7
-------
flows through the liquid surface and a demister. The demisting
pad prevents acid mist from exiting with the scrubbed gas and
provides a final hydrolysis reaction site for any EO remaining in
the gas stream. Inert gases (i.e., CFC-12 and CO2) are exhausted
unreacted to the atmosphere. After ethylene glycol builds up in
the stream to a maximum recommended level of 60 percent, the
scrubber liquor is neutralized and disposed or sold.17 (See
Section 4.1.1.1.3 for a more detailed discussion of waste
disposal.) Possible methods of determining the scrubbing liquor
changeout point include (1) using liquid level indicators,
(2) using specific gravity detectors, and (3) measuring the
amount of EO charged to the sterilizer. Reaction towers have
been installed for chambers ranging from 1.4 m3 (50 ft3) to 45 m3
(1,600 ft3).18'19 Seven of the 188 commercial sterilization
facilities represented in EPA's data base use reaction towers to
I
control EO emissions from sterilizer vents. The sterilizers at
these seven facilities range in volume from 4m3 (140 ft3) to
27 m3 (960 ft3). Annual ethylene oxide use at these seven
facilities ranges from 2.1 Mg (5,000 Ib) to 49 Mg
(110,000 Ib),2/3/5 ;
Manufacturers of reaction/detoxification towers claim
99+ percent EO removal efficiency by weight.18'20 Third-party
laboratory test results indicate that EO emission reductions
greater than 99.8 percent can be achieved with reaction towers.
However, a detailed review of the available test data indicates
that 99.0 percent is the highest EO removal efficiency .that can
be achieved by acid hydrolysis techniques on a continuous
basis.16
4.1.1.1.3 Waste disposal. The spent liquor from!acid
hydrolysis of EO is typically 40 to 60 weight percent ethylene
glycol and has a pH of 0.5 to 2.0. Because of the low pH, the
solution is considered a hazardous waste and thus, requires
special handling procedures for shipping. However, the spent
liquor can easily be neutralized with sodium hydroxide (caustic)
prior to disposal. Two recovery companies have been identified
that are willing to purchase the aqueous ethylene glycol
19
4-8
-------
solution.21'22 Both of these companies require that sodium
hydroxide be used for neutralization and will pick up the
solution at the sterilization facility. A third recovery company
will accept the spent scrubbing solution on a no cost/no payment
basis, except for shipping charges.23 Neutralized scrubbing
solution may also be disposed to a landfill or incinerator.
4.1.1.2 Oxidation. Two methods of oxidizing EO are
(1) thermal oxidation with flares and (2) catalytic oxidation
with a solid-phase catalyst.
4.1.1.2.1 Thermal oxidation. Ethylene oxide, which has a
high heating value, a relatively low ignition temperature, and a
very wide range of mixtures combustible in air (see Table 3-4),
can be easily and efficiently destroyed by thermal oxidation
using flares. Thermal oxidation of EO produces carbon dioxide
and water as follows:
2 C2H40+5 02
4 C02+ 4 H20
thermal oxidation
Three of the 188 commercial sterilization facilities
represented in the EPA data base reported using flares to control
EO emissions when pure EO was used as a sterilant gas.2'5 One of
these facilities has one 76.7-m3 (2,710-ft3) chamber and uses
26 Mg (57,000 Ib) of EO per year. Another facility has three
chambers ranging in size from 75.2 to 76.9 m3 (2,655 to
2,715 ft3) and one smaller 1.7-m3 (60-ft3) chamber; this facility
uses 89 Mg/yr (197,000 Ib/yr) of EO.2 The third facility has one
11-m3 (400-ft3) chamber and uses 25 Mg/yr (54,000 Ib/yr) of EO.5
Because of difficulties with sustaining combustion, commercially
available flares are not applicable for facilities emitting only
small amounts of EO.
A manufacturer of flare burners for the control of EO
emissions claims greater than 99-percent control efficiency for
pure EO, but no data were provided to substantiate this claim.24
The EPA's position is that flares operated within specified
4-9
-------
conditions of waste gas heat content and flare exit velocity will
achieve at least 98-percent destruction efficiency.25
Flares can also be used with EO/CO2 sterilant gasjmixtures
(e.g., 10/90) but are not designed for use with EO/CFcfl2
mixtures (e.g., 12/88).24/26 The EPA has not in the past and
does not now recommend the use of flares to control emission
streams containing halogenated compounds (e.g., CFC-12) because
i.
corrosive or toxic byproducts may form. As shown below, thermal
oxidation of CFC-12 may produce the following corrosive or toxic
byproducts at the high temperatures (400° to 800°C [800° to
1500°F]) associated with the use of flares:
CF2C12-K>2
CFC-12
COC12 Phosgene
thermal oxidation COF.
HCl'
HF
CF4
C12
CO
Carbonyl fluoride
Hydrogen chloride
Hydrogen fluoride
Carbon tetrafluoride
Chlorine |
Carbon monoxide
4.1.1.2.2 Catalytic oxidation. Catalytic oxidation of EO
occurs in the presence of a solid-phase catalyst as follows:
2 C2H40+5 02
4 CO2+4 H2O
catalytic oxidation
This control technique is applicable to pure EO, EO/CO2 mixtures,
and EO/CFC-12 mixtures. The CFC-12 does not react at the
temperatures (150° to 180°C [300° to 350°F]) that occur during
catalytic oxidation, and, therefore, the toxic CFC byproducts
that result from the higher temperatures associated with thermal
oxidation are not produced. During an EPA-sponsored test of a
catalytic oxidation unit, no CFC decomposition byproducts were
detected; the detection limit was 200 parts per billion (ppb) for
the analyte chloride ion. The maximum operating temperature of
the unit during testing was 155°C (311°F).27 !
A schematic of a catalytic oxidizer is shown in Figure 4-3.
The spent sterilizer gas is first mixed with a large volume of
air to reduce the control device inlet EO concentration to
5,000 parts per million by volume (ppmv) or less. This dilution
4-10
-------
I— O
CO -
CD
I
4J
(0
>i
(0
§
•H
•P
(0
T3
•H
X
o
o
•H
n
I
0)
a:
LU
O. CC <-}
S3 So
z —• <: o
5£
•—<
es
4-11
-------
prevents excessive catalyst bed temperatures (which can damage
the catalyst) from occurring during the oxidation of EOl The gas
stream passes through a filter for dust removal and then is
preheated to the reaction temperature with steam or electricity.
The gas then enters the catalyst bed(s), where the EO is
oxidized. Part of the exiting gas stream may be recycled for
heat recovery before being vented to the atmosphere. One
manufacturer also sells a catalytic oxidizer that uses excess
catalyst, instead of diluent air, to absorb the heat of;
oxidation.28 Because of cost considerations, the excess-catalyst
system has been used, thus far, only for chambers less than 1 m3
[40 ft3]) in volume.28 j
Two of the 188 commercial sterilization facilities!
represented in the EPA data base reported the use of a catalytic
oxidizer to control EO emissions from the chamber vent.2'4 One
facility has one 4-m3 (130-ft3) chamber and uses 0.5 Mgi/yr
(1,000 Ib/yr) of EO in an EO/C02 sterilant-gas mixture.P The
other facility has an 18-m3 (600-ft3) sterilizer and uses 7 Mg
(15,000 Ib) of EO annually; the EO concentration to the! control
A '
unit at this facility is regulated by throttles.*
Because catalytic oxidation is applicable to the -control of
low EO concentrations, facilities can manifold other EO; emission
sources (e.g., aeration chambers or rooms, sterilizer hjood and
door vent, and the gas cylinder room) to the control deyice. In
addition, if the catalytic oxidizer requires diluent air, these
low-concentration emission sources can provide part or jail of the
necessary diluent. !
Manufacturers of catalytic oxidation units claim EO
destruction efficiencies greater than 99.9 percent.29'3!0 Third-
party testing and an EPA-sponsored test support these claims for
small (<0.85 m3 [<30 ft3]) sterilizers.27'29 |
4.1.1.3 Condensation/Reclamation Systems. Recovery of
sterilant gas mixtures is possible using a reclamation system.
The sterilant gas mixtures will condense under conditions of
4-12
-------
reduced temperature and increased pressure, but precautions are
necessary to avoid explosions.
Figure 4-4 is a schematic of a condensation/reclamation
system for a 12/88 sterilant-gas mixture. (See Table 3-4 for
physical and chemical properties of CFC-12.) After each
sterilization cycle, the 12/88 gas is withdrawn and passed
through one of the two desiccant beds next to the chamber. (One
of the desiccant beds is regenerated while the other is in use.)
The dried 12/88 gas then passes to a compressor where it is
compressed to 345 kilopascals (65 psia) to improve condensation
efficiency.31 The compressed gas is piped to a separate
explosion-proof room, where it passes through a pressurized
condenser that is chilled by ethylene glycol to about -18°C
(0°F).31 The liquid 12/88 mixture is collected in a pressurized,
chilled holding tank. The noncondensed gas is recirculated to
the chamber and back through the condenser. The liquid collected
in the holding tank is transferred to a pressurized reblending
tank where the liquid is mixed and its composition determined by
infrared analysis. The liquid is then adjusted to the
12/88 (weight percent) ratio by adding the necessary amount of EO
or CFC-12. When the correct ratio is obtained, the liquid is
transferred to a pressurized storage tank in the chamber room.31
Although the reclamation.cycle could be continued
indefinitely, the amount of EO recovered declines to the point
where it is not cost effective to continue the reclamation cycle
after about three passes through the system (i.e., typically 60
to 90 minutes). The majority of the EO (80 to 85 percent) is
recovered during this time. Also, increasing the reclamation
time would require that products spend additional time in the
sterilizer and could affect the plant's operating schedule.
However, even if the reclamation time was increased, this system
is not designed for low EO concentrations. Therefore, if this
type of control system is used, add-on controls (e.g., catalytic
oxidation or a small scrubber) need to be considered for the EO
remaining in the chamber after the reclamation cycle is complete.
4-13
-------
Q>
4J
M
>i
M
(0
E
(0
•H
O
Q)
(0
in
c
0)
0)
4-14
-------
Six of the 188 commercial sterilization facilities
represented in the EPA data base reported the use of
condensation/reclamation systems; one of these facilities
reported an 83-percent EO recovery efficiency, four reported
80 percent, and one reported 50 percent.2'3'5 These six
facilities recover EO/CO2 and EO/CFC-12 sterilant gases. Five of
these facilities use over 23 Mg/yr (50,000 Ib/yr) of EO.2'3'5
The sixth facility uses just over 6.8 Mg/yr (15,000 Ib/yr).2'3
The chamber sizes range from 5 to 45 m3 (190 to 1,580 ft3) at
these six facilities.2'3'5
The condensation/reclamation systems currently available are
designed for the high volumetric flow rates of larger,
industrial-sized chambers. The systems are not technically or
economically feasible for use with smaller chambers or at
facilities that use small amounts of EO.
4.1.1.4 Gas/Solid Reactor. A fourth control technique to
control vent emissions is a dry, solid-phase system that
chemically converts EO and then binds the product to the solid
packing.1 This system is generally paired with an acid water
scrubber. The system operates at room temperature. There are no
liquid waste streams produced; the solid waste is returned to the
vendor for recycling.32'33 Although the gas/solid reactor can
handle high EO concentrations (i.e., >100,000 ppmv) for brief
periods of time, it is designed for low concentrations such as
the exhaust from an acid-water scrubber. The manufacturer of
this device markets a two-stage control system, which consists of
an acid-water scrubber and the gas/solid reactor. (The company
also sells the stages separately.) The majority of the EO is
removed by the scrubber, which is specifically designed for small
sterilizers (<2 m3 [70 ft3]). The gas/solid reactor removes the
residual EO exiting the scrubber and, because it is designed for
low EO concentrations, can also be manifolded to other emission
sources (e.g., aeration chambers, sterilizer hood and door, and
gas cylinder storage room).
The manufacturer of this system claims greater than
99.9 percent efficiency for the gas/solid reactor.1 However,
4-15
-------
this efficiency is based on a test performed with an inlet EO
concentration of 140,000 ppmv, which is much higher than the
concentration of the scrubber outlet stream. In another test, no
EO was detected (with a lower detection limit of 0.1 ppmv) in the
gas/solid reactor outlet stream when the inlet stream (I.e.,
scrubber outlet stream) was 2 ppmv EO.32 Because of the inherent
problems associated with measuring low EO concentrations, the
actual efficiency of the system under normal operating conditions
presently cannot be determined. (See Section 4.1.3 for a more
detailed discussion of measuring low EO concentrations.)
However, the maximum removal efficiency that the gas/solid
reactor can achieve on a continuous basis is assumed to be
99.0 percent.
4.1.2 Sterilization Chamber Vacuum Pump Drain Emissions
I
Ethylene oxide drain emissions result from the use of vacuum
pumps that use once-through water as the working fluid.! Ethylene
oxide is infinitely soluble in water, and, therefore, a portion
of the EO evacuated from the chamber enters the drain w;ith the
vacuum pump water (see Figure 4-5a). The EO that enters the
drain with the vacuum pump water is subsequently released
uncontrolled from the air break in the drain line, sewer lines,
or the waste water treatment plant.6"8
The EO drain emissions can be eliminated by replacing the
existing once-through vacuum pump with a closed-loop j
(recirculating-fluid) vacuum pump. The recirculating fluid
(sealant) can be water, oil, or ethylene glycol.34 In this
closed-loop system, the water or liquid from the liquid-gas
separator is cooled in a heat exchanger and recirculated through
j
the vacuum pump (see Figure 4-5b). Because ethylene oxide is not
soluble in oil or ethylene glycol and will off-gas from water as
it is recirculated, nearly all of the EO will be emitted through
the liquid-gas separator (chamber) vent. (Techniques for control
of chamber vent emissions are discussed above.) In addition,
mechanical seals are used to eliminate leakage (and, thus, any EO
emissions) from the pump.35
4-16
-------
VACUUM
PUMP
EVACUATED GAS FROM
STERILIZATION CHAMBER
£0 VENT
EMISSIONS
LIQUID-GAS
SEPARATOR
WATER
VACUUM PUMP WATES
£0 DRAIN EMISSIO'
Figure 4-5a. Once-through liquid-ring vacuum pump.
VACUUM
' PUMP
EVACUATED GAS FROM
STERILIZATION CHAMBER
RECIRCULATED
WATER
EO VENT
EMISSIONS
LIOUIO-GAS
SEPARATOR
VACUUM PUMP
WATER
Figure 4-5b. Recirculating liquid-ring vacuum pump.
4-17
-------
Because the sterilization cycle operates under humid
conditions, some water will be condensed in the liquid-|gas
separator and, thus, mix with the liquid sealant in the pump. An
overflow collection tank is used to maintain a constant amount of
sealant recirculating in the pump.35 If ethylene glycol is used
as the sealant, the contaminated glycol will eventually need to
be disposed and replaced with a fresh charge.35 Howevejr, if oil
is used as the sealant, the condensed water can be drained off
the bottom with minimal oil loss because of the immiscibility of
oil and water.34
4.1.3 Chamber Exhaust Emissions
Some facilities, in an effort to meet the EO permissible
worker exposure level set by the Occupational Safety and Health
Administration (OSHA) (see Section 3.5), have installed chamber
exhaust vents on sterilizers. The purpose of the exhaust is to
quickly dilute EO concentrations in the sterilizer chamber void
space and thereby prevent exceedances of exposure limits for
workers. Other-facilities have installed hoods above the chamber
door to reduce worker exposure. At present, there are [no data on
controlled chamber exhaust emissions in the commercial 'sterili-
zation data base. As with aeration rooms (see Section 4.1.4.1
for a more detailed discussion), the low-concentration, high-
flow-rate exhaust streams of chamber exhausts limit the!
feasibility and efficiency of add-on controls (particularly
thermal oxidation and condensation/reclamation), and the lower
i
detection limit of most analytical methods may make it !impossible
to determine the efficiency of the control devices at tine
concentrations typical of the chamber exhaust stream. [The same
control techniques that are applicable to the control Qf aeration
room emissions (i.e., catalytic oxidation units and gas/solid
reactors) may also be applicable to chamber exhaust emissions.
In addition, acid/water scrubbers may also be feasible.!
The typical chamber exhaust provides a flow rate of
84 m3/min (3,000 ft3/min), and the EO concentration in Ithe void
volume as the sterilizer door is opened is estimated to; be 500 to
15,000 ppmv, depending on sterilizer operating parameters. As
4-18
-------
the diluent air enters the chamber, the EO concentration rapidly
decreases to below the 1 ppmv OSHA limit. Some EO may evolve
from the product during the chamber exhaust cycle, but this
amount is negligible when compared to the EO concentration in the
void volume.
4.1.3.1 Acid-Water Scrubber. As discussed in
Section 4.1.1.1, acid-water scrubbers are commonly used to
control sterilizer vent emissions, which have low to moderate
flow rates (0.7 to 14 m3/min [25 to 500 ft3]) and potentially
high EO concentrations (400,000 ppm).20 Under these conditions,
acid-water scrubbers can achieve EO removal efficiencies of
99 percent or greater.6 While it is technically feasible to
control the higher-flow-rate, lower-concentration emissions from
the chamber exhaust, the EO removal efficiency of the scrubber
may be reduced. The potential decrease in efficiency would be
due to decreased residence times of EO in the scrubber and
because there would be less EO to react in the scrubbing liquor.
Because an acid-water scrubber has never been demonstrated to
i
control chamber exhaust emissions, the control efficiency of the
unit under these conditions is unknown.
4.1.3.2 Catalytic Oxidation System. As discussed in
Section 4.1.1.2.2, the inlet EO concentration for excess-air
catalytic oxidizers is typically reduced to 5,000 ppmv or less,
which means that use of this control device could be feasible
with chamber exhaust streams. Catalytic oxidation units are also
commercially available to handle flow rates from chamber
exhausts.37'38 While no catalytic oxidizers are known to have
been installed to control chamber exhaust emissions, units have
been installed to control aeration room emissions. As discussed
r
in Section 4.1.4.1, the control efficiency of these units has not
been adequately demonstrated.
4.1.3.3 Gas/Solid Reactor. Like the catalytic oxidizers,
the gas/solid reactor is designed for low concentration
(<100 ppm) inlet streams.33 As discussed in Section 4.1.4.1,
gas/solid reactors are being used for flow rates up to 42 m3/min
(1,500 ft3/min), and systems can be sized to accommodate larger
4-19
-------
flow rates.39 However, the control efficiency of these units
with low EO concentrations has not been adequately Determined.
4.1.4 Aeration Room Vent Emissions i
4.1.4.1 Aeration Rooms. Most commercial sterilization
facilities aerate the sterile products in large, warehouse-type
aeration rooms that are typically 280 to 2,800 m3 (10,000 to
100,000 ft3) in volume but may be larger than 14,000 m3
(500,000 ft3).4 The ventilation rates are, generally, in the
range of 112 to 560 m3/min (4,000 to 20,000 ft3/min) but may be
as high as 1,680 m3/min (60,000 ft3/min).4 These large flow
rates are necessary to maintain a low EO concentration ;in the
room to comply with OSHA standards (see Section 3.5). 'Data from
a cross-sectional survey (44 facilities) of the 188 commercial
sterilization facilities represented in EPA's data base indicated
an average 8-hour, time-weighted average (TWA) EO concentration
of 2.5 to 3 ppmv in aeration rooms.4 i
Two issues of concern regarding the control of aeration room
emissions are: (1) most EO emission control devices are
impracticable for the low-concentration, high-flow-rate exhaust
i
streams from aeration rooms; and (2) the lower detection limit of
most analytical methods may make it impossible to determine the
true control efficiency of the low EO concentrations (less than
1 ppmv) found in most aeration rooms. Hydrolysis, therfmal
oxidation, and condensation/reclamation presently have tnot been
demonstrated to be practicable control techniques for low-
concentration, high-flow-rate gas streams. However, catalytic
oxidation and the gas/solid reactor system have the potential to
control aeration room emissions. Catalytic oxidation units are
commercially available to handle flow rates from less tihan
1 m3/min (40 ft3/min) to approximately 340 m3/min
(12,000 ft3/min),37'38 The catalytic oxidizers are modular, and
systems can be designed to handle higher flow rates; however, the
increased size of the system for high flow rates can restrict its
practical use. Gas/solid reactors are being used for flow rates
up to 42 m3/min (1,500 ft3/min), and systems can be designed to
4-20
-------
handle any flow rate; however, as with catalytic oxidation, the
system size can become impractical.39
The manufacturers of catalytic oxidizers and gas/solid
reactors claim EO destruction efficiencies greater than
99.9 percent and offer the results of third-party tests to
support these claims.32'40'41 However, test data are not
available on the efficiencies of the control units operating
under conditions that are typical of aeration room exhaust
streams (i.e., low concentrations and high flow rates).
Generally, the control units are tested by sending the
'control device a stream of EO with a much higher concentration
(e.g., 100 to 140,000 ppmv) than that associated with normal
operating conditions.32'40'41 The results of these tests are the
efficiencies reported by the manufacturers. However, these test
results may be misleading because: (l) it has not been
demonstrated whether the control units perform at the same
efficiencies under normal operating conditions (i.e., very low
inlet concentrations) as during test conditions (i.e., controlled
flow, high concentration); (2) EPA has not conducted its own
emission test program to verify the efficiency of control devices
with aeration room emissions; and (3) EPA has not yet developed
or approved a test reference method that is applicable' to the
evaluation of the efficiencies of these control devices with
aeration room emissions.
The lower detection limits of most analytical procedures
that are used to measure EO concentrations are approximately
0.5 ppmv to 1 ppmv, which is equal to or greater than the EO
concentrations in many aeration rooms. Although one testing
laboratory reportedly used a method with a detection limit less
than 0.1 ppmv, the test data have not been verified by EPA, and
it is unknown whether this method can be applied to high flow
rates. (The flow rate tested was 14 m3/min [500 ft3/min].)41
Also, because of the reactivity of EO, the validity of detection
limits below l ppmv, and particularly below 0.5 ppmv, is
questionable.42 Because the detection limits of the analytical
methods (in ppmv) are so close to the room concentrations,
4-21
-------
testing under normal operating conditions may yield an 'efficiency
that can only be calculated to be equal to 50 percent or less.
Three possible techniques for reducing EO emissions from
aeration rooms are to (1) recirculate the air from the aeration
room control device to the aeration room, (2) replace the
warehouse-type aeration rooms with smaller, heated aeration
chambers, or (3) modify the evacuation arid air wash pha|se of the
sterilization cycle. The first two techniques increase the EO
concentration in the aeration room and lower the flow rate, which
makes both control of the emissions and testing of the |control
efficiency more practical. The third alternative lowers the EO
•
emissions from the aeration room by decreasing the residual EO in
the product prior to aeration. These techniques are discussed in
more detail below.
The first alternative refers to routing the aeration room
air through an emission control device and back to the iaeration
room. A small amount of makeup air is added to the control
device exit stream to regulate the room temperature. This
practice increases the room temperature and, therefore, increases
the diffusion rate of EO from the product, producing a ;higher EO
concentration in the room. (Worker exposure and compliance with
the OSHA standards will need to be considered if frequent worker
access to the room is required.) Catalytic oxidation a|nd the
gas/solid reactor are more applicable to gas streams having
i
increased EO concentrations and decreased flow rates associated
with this process than to typical aeration room emissions. In
addition, increasing the room temperature reduces the energy
costs of preheating the inlet stream to the catalytic oxidizer.
Hydrolysis, thermal oxidation, and condensation/reclamation are
not applicable control techniques because the EO concentrations
i
are too low (<20 ppmv) for these techniques to be practicable.
Because the room air is recirculated and not vented to jthe
atmosphere, this technique eliminates practically all aeration
room emissions; only a small amount of the emissions from the
control device are vented to allow fresh makeup air to ienter the
room. This practice of recirculating the aeration room air is
4-22
-------
used toy 2 of the 188 commercial sterilization facilities
represented in the EPA data base.3 The aeration rooms at these
two commercial sterilization facilities are each approximately
140 m3 (5,000 ft3) in volume.4 These two facilities manufacture
synthetic rubber products, which retain a large amount of
residual EO and, therefore, require a longer aeration period than
the majority of products that are sterilized with EO. The
facilities installed the recirculating system to decrease the
aeration time and the residual EO concentrations in the
products.41 A catalytic oxidation system is used to control the
EO emissions at these facilities and to provide hot air to heat
the room.43
Another alternative is to replace the large, warehouse-type
aeration rooms with smaller (70 m3 [2,500 ft3] or less), heated
aeration chambers and control the emissions from the chamber. In
this process, instead of storing the sterile products in a
warehouse and aerating at normal room temperatures, the products
are aerated in heated (>43°C [110°F]), insulated chambers. The
emissions from these units can be controlled by catalytic
oxidation or the gas/solid reactor system. Emissions from the
control device can be recirculated to the aeration chamber or
vented to the atmosphere. The aeration chambers can be filled
approximately 40 to 75 percent full and still allow sufficient
air space for off-gassing.28'44 The aeration chamber is heated
with either supplemental heat or hot air from the control device
if catalytic oxidation is used. Several commercial sterilization
facilities, particularly contract sterilizers, are aerating at
least some of the sterile products in heated, aeration
chambers.4'38 Structures used for aeration include insulated
shipping containers, modified walk-in coolers (which are heated
instead of cooled), and manufactured units designed specifically
for the heated aeration process.45"47 Most of these facilities
have installed these chambers to reduce the aeration time or the
residual EO concentration in the products. The heated aeration
chambers are similar to the first technique described above
(i.e., the practice of recirculating the aeration room air) in
4-23
-------
that the EO concentration in the chamber will increase as a
result of EO off-gassing from the product due to elevated
temperature. |
Another strategy for reducing aeration room emissions is to
modify the evacuation/air wash phase of the sterilization cycle.
Residual EO in the product can be reduced by performing I
additional sterilization chamber purges. However, this procedure
requires that products be held for additional time in the
sterilizer and could affect plant operating schedules. JThe
potential reduction in residual EO with evacuation-phase
modifications is product dependent. Results from tests|performed
at one facility that fumigates spices showed an average|reduction
in residual EO of 26 percent for four different spices following
evacuation-phase modifications.15 Some facilities aerate in the
sterilizer, with and without cycle modifications.4 Aeration
emissions from the sterilizer can be sent to the sterilizer
control device. However, the removal efficiencies of the
hydrolysis techniques have not been determined for the low inlet
concentrations associated with aeration emissions. Also,
condensation/reclamation would not be practicable for controlling
these low concentrations.
An additional system currently used to control EO emissions
from aeration rooms is an acid-impregnated carbon adsorption
system. When such a system is used, emissions from an aeration
room or aeration cabinet (see Chapter 4.1.4.2) are ducted to the
carbon adsorption system which typically consists of approxi-
mately 20 acid-treated carbon trays. The carbon in thebe carbon
i
trays has been treated with a strong acid (e.g., sulfuric acid)
and is humidified. The EO in the emissions from the aeration
room(s)/cabinet(s) is hydrolyzed to ethylene glycol. Because of
its increased affinity for EO, the EO removal efficiency of the
acid-impregnated carbon greatly exceeds that of plain cjarbon.
However, the actual removal efficiency for this emissions control
device have not yet been determined.48 ;
4.1.4.2 Aeration Cabinets. Some commercial sterilization
facilities use aeration cabinets instead of aeration rojoms.
4-24 ;•
-------
These cabinets are similar in appearance and size (
-------
fumigate beehives. However, one State Department of Agriculture
is working on the development of an acid-water scrubber for
portable fumigation units.3 !
4.3 ALTERNATIVES TO EO STERILIZATION
In some cases, radiation sterilization can replace EO
sterilization. Radiation sterilization is used for abput half of
the products sterilized in the United States.50 However, not all
products can be sterilized with radiation; plastics can become
broken, discolored, or rendered malodorous, and Teflon® and
acetyl delrin are damaged by radiation.50'51 According to
industry representatives, most of the commonly used plastics have
been or are in the process of being reformulated to withstand
radiation.52'53 Therefore, the potential use of this alternative
may increase. !
There are several chemical alternatives to EO sterilization
(e.g., chlorine dioxide, gas plasma, hydrogen peroxide,; and
ozone). However, these chemicals do not necessarily offer
environmental improvements over EO. Other alternatives include
X-ray sterilization (a new, developing technology), deep freezing
I
(museum and spice industry), and increased use of disposable
medical items in hospitals. However, none of these alternatives
can replace the use of EO for all applications, and they may have
adverse environmental impacts as well. For example, the
i
increased use of disposables may conflict with a pollution
prevention program. Additionally, there may be significant
health effects if these alternatives produce less-effective
sterilants
54
4.4 RETROFIT CONSIDERATIONS
All of the control devices discussed above can be ,
retrofitted to existing EO bulk sterilization chambers.; However,
the use of flares in urban areas is prohibited because of safety
hazards. There are no retrofit problems associated wit^i the
replacement of once-through vacuum pumps with closed-loop
recirculating vacuum pumps for control of drain emissions.
4-26
-------
4.5 IMPACTS OF CFC REGULATION ON EO EMISSION CONTROLS
Federal regulations for stratospheric ozone-depleting
chlorofluorocarbons (CFC's) have been developed under EPA's
Stratospheric Ozone Protection Program (SOPP). The use of CFC's
in sterilant gases is one of the source categories subject to the
CFC regulation. The most popular sterilant gas mixture, 12/88,
contains 88 percent by weight CFC-12, which is an ozone-depleting
CFC. Seventy-five percent of the 188 commercial sterilization
facilities represented in the EPA data base use 12/88 at least
part of the time.2'3'5 The requirements of a CFC regulation
would not change the ability of a sterilization facility to
control EO emissions. The explosion-proof
condensation/reclamation system discussed above recovers CFC-12
emissions in addition to EO emissions. However, if this control
device is used, add-on controls (e.g., catalytic oxidation or a
small scrubber) need to be considered for the EO remaining in the
chamber after the reclamation cycle is complete. Also, a
dedicated CFC-12 condensation/reclamation system that follows the
acid-water scrubbing of EO to ethylene glycol would not have to
be explosion-proof.55 Some facilities may switch to sterilant
gases that do not contain CFC-12 (such as 10/90 and pure EO), in
which case the EO control techniques discussed above would still
be applicable.
4.6 REFERENCES FOR CHAPTER 4
1.
2.
3.
Safe-Cell™ product brochure. Attachment to letter from
Kruse, R., Advanced Air Technologies, Inc., to J. Farmer
EPA. May 31, 1988.
Letter and enclosures from Jorkasky J. Health Industry
Manufacturer's Association (HIMA), to D. Markwordt, EPA:CPB.
February 21, 1986. Survey responses from HIMA members.
Memorandum. deOlloqui, V., MRI, to Project Files. June
1990. List of respondents to July 1986 Section 114
information request regarding the use of ethylene oxide by
miscellaneous sterilization and fumigation facilities.
4-27
-------
8,
10,
11,
12,
13
14,
15,
16,
Memorandum. deOlloqui, V., MRI, to Project Files. June
1990. List of respondents (44 facilities) to July 1988 EPA
Section 114 letter regarding sterilizer operating
parameters, existing controls, vacuum pumps, and aeration
rooms. 1
i.
Memorandum. deOlloqui, V., MRI, to Project Files.! June
1990. List of respondents (39 facilities) to July 1989 EPA
Section 114 letter to facilities with maximum individual
risks greater than 10~3. i
I
Ethylene Oxide Product Information Bulletin. Unidn Carbide
Corp., Ethylene Oxide/Glycol Division. 1983. |
Letter from Buonicore, A., Chemrox, Inc., to D. Markwordt.
EPA:CPB. August 27, 1984. Comments on sources o^ ethylene
oxide emissions draft report. !
Conway, R., Waggy G., Spiegel M., and R. Berglund.
Environmental Fate and Effects of Ethylene Oxide, i
Environmental Science and Technology. 17(2);107-I12.
1983
Questionnaire for Croll-Reynolds Ethylene Oxide Scrubber—
Customer Specifications. Croll-Reynolds Company, i
Westfield, NJ. October 1985.
i
Telecon. Newton, D., MRI, with T. Urban. Chemrox, Inc.
February 13, 1986. Discussion about disposal of scrubber
liquor containing ethylene glycol. i
I
Newsletter about EO control. Chemrox, Inc., Bridgeport, CT.
Volume 1, No. 1. October 1983.
Certification Testing Report.
Chemrox Inc., Bridgeport, CT.
BCA Project No. 85-1260.
October 29, 1985.
Letter from Desai, P., Chemrox, Inc., to S. Wyatt.i EPA:CPB.
September 17, 1987. Comments on draft BID Chaptei^ 4 for
ethylene oxide NESHAP.
Sampling/Analytical Method Evaluation for Ethylene' Oxide
Emission and Control Unit Efficiency Determinations. Final
Report. Radian Corporation. Research Triangle Park, NC.
April 5, 1988. !
Desai, P. Performance Test Report: DEOXX™ Ethylene Oxide
Detoxification System. Chemrox Project No. 85-260|. October
1985.
i
Memorandum. Srebro, S., MRI, to D. Markwordt. EP|A:CPB.
October 5, 1988. Examination of Ethylene Oxide Control
Efficiencies. !
4-28
-------
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
Memorandum. Srebro, S., MRI, to D. Markwordt. EPArCPB.
March 20, 1987. Capital cost, annual!zed cost, and cost
effectiveness of reducing ethylene oxide emissions at
commercial sterilization facilities. 80 p.
Product Data Sheet. Environmental Tectonics Corporation.
Enclosure to letter from Peters, J., Environmental Tectonics
Corporation, to B. Nicholson. MRI. June 10, 1987.
Letter and attachments from Smith, D., Damas Corp., to
S. Wyatt. EPAiCPB. September 21, 1987. Comments on draft
BID Chapter 4 for ethylene oxide NESHAP.
Meeting Minutes. Beall, C., MRI, to D. Markwordt. EPArCPB.
April 30, 1986. Damas Corp. and Johnson & Johnson. 9 p.
Telecon. Srebro, S., MRI, with J. Hoffman. Med-Chem
Reclamation, Inc. March 6, 1989. Discussion about recovery
of ethylene glycol from ethylene oxide scrubbing liquor.
Telecon. Srebro, S., MRI, with J. Duvow. Chemstrearns,
Incorporated. March 16, 1989. Discussion about recovery of
ethylene glycol from ethylene oxide scrubbing liquor.
Telecon. Srebro, S., MRI, with K. Dalton. High Valley
Chemicals. March 16, 1989. Discussion about recovery of
ethylene glycol from ethylene oxide scrubbing liquor.
Letter and attachments from Smith, S., John Zink Company, to
B. Coronna. MRI. October 3, 1986. Information about the
John Zink EO flare.
Flare Efficiency Study. U. S. Environmental Protection
Agency. Research Triangle Park, N.C. Publication No. EPA-
600/2-83-052. July 1983. p. 5.
Telecon. Soltis, V., MRI, with B. Duck. John Zink Company.
July 8, 1987. Discussion about EO sterilant gas mixtures
and the use of flares.
Meiners, A. (MRI) Ethylene Oxide Control Technology
Development for Hospital Sterilizers. Prepared for U.
Environmental Protection Agency. Publication No.
EPA-600/2-88-028. May 1988.
S.
Telecon. Srebro, S., MRI, with D. Meo. DM3 Incorporated.
January 13, 1989. Discussion about CATCON catalytic
oxidation systems.
Letter and attachments from Olson, C., Donaldson Company,
Inc., to D. Markwordt. EPA:CPB. March 21, 1988. Test data
for EtO Abator™ for sterilizer chamber emissions.
4-29
-------
30. Telecon. Srebro, S., MRI, with D. Meo. DM3 Incorporated.
December 2, 1988. Discussion about the CATCON catalytic
oxidation systems. !
31. Memorandum. Beall, C., MRI, to p. Markwordt. EPA|:CPB.
Trip Report: Sterilization Services of Tennessee., Memphis,
Tennessee, on March 18, 1986. '
32. Letter and attachment from Hammer, D., Advanced Air
Technologies, (Michigan Science and Engineering) to
D. Markwordt. EPArCPB. June 22, 1988. Transmitting test
data for the Safe-Cell™ system. i
i
33. Telecon. Srebro, S., MRI, with D. Hammer. Advanced Air
Technologies. December 11, 1989. Discussion about design
of gas/solid reactor.
34. Buonicore, A. (Memorox, Inc.) In-Plant Programs jto Reduce
Ethylene Oxide Worker Exposure Levels. Chemrox, ijnc.
Bridgeport, CT. August 1984. i
35. EO-VAC™ Closed Loop Vacuum Product Information Shejet.
Chemrox, Inc. Bridgeport, CT. May 1987.
36. Memorandum. deOlloqui, V., MRI, to D. Markwordt. EPA:CPB.
April 1990. Costing of Control Alternatives for t^he Rear
Chamber Exhaust.
37. Telecon. Nicholson, R., MRI, with C. Olson. Dona'ldson
Company, Inc. May 12, 1988, and June 13, 1988. Discussion
about applicability of catalytic oxidation to large aeration
rooms and location of facilities using the EtO Abater™.
38. Letter and attachments from Meo, D., DM3 Incorporated, to
S. Srebro. MRI. January 13, 1989. Transmitting j
information about the CATCON system.
I
39. Letter and attachments from Kruse, R., Advanced Air
Technologies, to R. Nicholson. MRI. June 15, 1988.
Transmitting information about Safe-Cell™ system.
40. Letter and attachments from Olson, C., Donaldson Company, to
S. Srebro. MRI. November 9, 1988. Transmitting |test data
for the EtO Abator™.
41. Report of Air Pollution Source Testing Conducted at Isolab
Corporation. Engineering Science, Incorporated. [Pasadena,
CA. September 27, 1988. Attachment to letter from Meo, D.,
DM3, Incorporated, to S. Srebro. MRI. December 16, 1988.
42. Analytical Chemistry.
24-60.
Volume 60. 1988. pp. 24-54 to
4-30
-------
43. Telecon. Shine, B., MRI, with L. Outright. Seamless
Hospital Products. October 20, 1988. Discussion about the
catalytic oxidizer used to control aeration room emissions
at Seamless.
44. Telecon. Friedman, E., MRI, with R. Shumway. Medtronic,
Incorporated. January 18, 1989. Discussion about the
heated aeration cells at Medtronic.
45. Memorandum. Srebro, S., MRI, to D. Markwordt. EPA:CPB.
Site visit: Medtronic, Incorporated. Anaheim, CA. on
December 9, 1988.
46. Memorandum. Srebro, S., MRI, to D. Markwordt. EPA:CPB.
Site visit: lolab, Incorporated, Claremont, CA. on
December 9, 1988.
47. Product brochure. Chemrox, Incorporated. Hot Cell™ heated
aeration unit.
48. Trip Report. Srebo, S., MRI, to D. Markwordt. EPA:CPB.
Medtronic, Inc., Anaheim, CA. on December 9, 1988.
49. Contact report. Srebro, S., MRI, with D. Meo
DM3 Incorporated. December 9, 1988. Discussion about
excess-catalyst catalytic oxidation system.
50. Telecon. Soltis V., MRI, with J. Jorkasky. Health Industry
Manufacturers Association. March 2, 1987. Discussion about
trends in the sterilization industry.
51. Telecon. Beall, C., MRI, with A. Chin. Radiation
Sterilizers, Inc. February 22, 1986. Discussion about
gamma radiation.
52. Chin, A. (Radiation Sterilizers Incorporated). Cobalt
Growth from 1978 to 1988. Presented at the Health Industry
Manufacturers' Association (HIMA) Sterilization in the
1990's conference. Washington, DC. October 31, 1988.
53. Apostolou, S., POLY-FOCUS. Radiation and Plastic: Friend
or Foe. Presented at HIMA conference. October 31, 1988.
54. Minutes. National Air Pollution Control Techniques Advisory
Committee: Ethylene Oxide Commercial Sterilizers, Status
Report. January 29-31, 1991, EPA-ESD.
55. Telecon. Srebro, S., with P. Desai. Chemrox, Incorporated.
January 20, 1987. Discussion about the FREOXX™ CFC-12
reclamation system.
4-31
-------
-------
5.0 REGULATORY ALTERNATIVES
5.1 INTRODUCTION
The purpose of this chapter is to present the regulatory
alternatives developed by EPA. These regulatory alternatives
represent the various courses of action that EPA could take in
controlling ethylene oxide emissions from commercial
sterilization facilities. The environmental, cost, and economic
impacts associated with the application of these alternatives to
this source category are presented in subsequent chapters. The
EPA has developed a facility-specific data base for this source
category; the impacts of the regulatory alternatives are based on
actual facility data.
Data in the EPA commercial sterilization (CS) data base
include facility-specific information from 196 facilities. This
information was obtained through a November 1985 Health Industry
Manufacturers' Association (HIMA) survey of medical equipment
suppliers and a July 1986 Section 114 information collection
request submitted to miscellaneous sterilizers and fumigators.
Additional information that was used to update the 1986 EPA CS
data base was obtained from two subsequent Section 114
information requests (July 1988 and July 1989). Facilities using
single-item sterilization (Sterijet®) units and beehive
fumigators used by State agriculture departments were not
included in the regulatory analyses because there are no
demonstrated control technologies for facilities using these
sterilization processes (see Section 4.2). Therefore,
188 facilities that operate a total of 404 sterilization chambers
are included in the regulatory alternatives analyses. These
5-1
-------
facilities are grouped by Standard Industrial Classification
(SIC) code into the following categories:
1. Medical equipment suppliers;
2. Pharmaceuticals; '
3. Other health-related industries;
4. Spice manufacturers;
5. Contract sterilizers;
6. Libraries, museums, and archives; and
7. Laboratories (research, testing, and animal breeding).
The data base includes information on the number of
chambers, chamber size, sterilant gas type, total annua|l
sterilant gas throughput, annual EO use, and levels of [emission
control.
5.2 REGULATORY ALTERNATIVES
I
Table 5-1 presents the regulatory alternatives EPA developed
to evaluate the environmental and cost impacts of potential
emissions controls on commercial sterilization facilities. The
regulatory alternatives represent incremental increases; in the
use of control devices and decreases in the EO use cutoffs that
are applicable to the emissions sources. The control devices
examined exhibited control efficiencies consistent with those
comprising the (MACT) floor. In other words, these devices
provide an emission reduction that is at least as stringent as
the average emission limitation achieved by the best performing
12 percent of the existing sources. The cutoff levels are based
on a facility's total annual EO use. Any facility with an annual
EO use rate greater than or equal to the particular cutoff level
would be subject to regulation. The analysis of emissions
control versus the impacts of control yielded a nonclustered,
continuous curve from which clear regulatory cutoffs were not
readily determined. However, the trend of the data indicates
that lower ethylene oxide annual use rates resulted in|higher
costs. The cutoff levels presented in Table 5-1 reflect this
trend. Sterilization chamber size was also considered as a basis
for regulatory alternatives. However, although the sterilization
5-2
-------
•a! 8.2
I g'g *
HII-
ON
O\
a
ON
QO
ON
M
eo
w
w
H
EH
a
8^
§
tn
1
00
1
00
S
«
g;
£
u
"p S
S-
ON
ON
o
H
I
in
se
o
S
s
w
oa
8
I £~
ill
i
ON
ON
1
oo
E
60
.tt
f
g
8
§
CS
00
5-3
-------
chamber size is related to the quantity of EO used, actual EO use
is a more direct measure of emissions. '
Regarding the 900 kilograms per year (kg/yr) (2,000 pounds
per year [lb/yr]) EO use cutoff for sterilizer vent and chamber
exhaust emissions controls, no plants in the CS data base using
less than 900 kg/yr (2,000 lb/yr) of EO have controlled; emissions
from sterilizer vents or chamber exhausts. Regarding the
18,160 kg/yr (40,000 lb/yr) EO use cutoff for aeration iroom
emissions control, none of the existing sources presently control
emissions from facilities using less than 18,160 kg/yr (40,000
lb/yr) of EO. Additionally, risk modeling data indicated that
emissions from sources above these cutoffs would pose aj more
significant threat to human health and the environment. Also,
cost estimates show that the cost impacts would be unreasonably
high for sources below these cutoffs. \
Estimates for emission rates and the cost of regulatory
compliance are based on the following control technologies: EO
emissions from the sterilizer vent would be controlled by an
acid-water scrubber, and vacuum-pump drain emissions wolild be
controlled by replacing the once-through, water-sealed,jvacuum
pump with a vacuum pump that has a closed-loop recirculation
system. All EO entering the vacuum pump would be routeci to the
control device through the sterilizer vent rather than being
split between the vent and drain, thus eliminating EO emissions
from the drain. Aeration room emissions would be controlled by
either a catalytic oxidizer or solid-bed reactor. The same
controls as the aeration room would apply to chamber exhaust
» S
emissions.
The alternatives are presented in decreasing order!of
stringency. Regulatory Alternative A represents the maximum
nationwide level of control. At this level, an estimated
99 percent of ethylene oxide emissions from commercial |
sterilization operations would be captured and controlled. All
facilities in the data base would be subject to the control
requirements because there are no facilities in the EPA data base
that use less than 11 kg (25 Ib) of EO per year. Regulatory
5-4
-------
Alternative B has less stringent maximum EO use cutoffs. Under
Regulatory Alternative C, a cap on chamber exhaust emissions
limiting emissions to baseline emission levels would be required;
additional controls are assumed necessary to maintain the
baseline level. Under Regulatory Alternatives D and E, the only
emission source to be controlled is the main sterilizer vent.
Regulatory Alternative E represents the MACT floor for existing
sources; at least 12 percent of the best performing existing
sources already apply a 99-percent efficient control device to
control sterilizer vent emissions.
It is important to note that while the efficiency of
acid-water scrubbers (at least 99-percent) is widely accepted,
the efficiencies assumed for the controls for aeration rooms and
chamber exhaust vents are not well supported. For purposes of
the analyses in this document, 99 and 98 percent were selected
for the aeration room and chamber exhaust vent control device
efficiencies, respectively. However, actual control efficiencies
may be lower given the high-flow, low concentration emissions
streams that are typical of these emissions sources.
5-5
-------
-------
6.0 ENVIRONMENTAL IMPACTS
This chapter presents estimated primary and secondary
impacts on air, water, solid waste, and energy for each of the
regulatory alternatives discussed in Chapter 5. Both beneficial
and adverse impacts, as well as potential emission reductions,
are assessed for the 188 facilities represented in EPA's 1989
commercial sterilization (CS) data base. Because no significant
growth is expected for this industry, the 5-year impacts are the
same as current impacts, and, therefore, only current impacts are
presented in this section.1'2
6.1 AIR POLLUTION IMPACTS
6.1.1 Baseline Emissions and Emission Reduction
Based on facility-specific data in the EPA 1989 CS data
base, baseline EO emissions and potential emission reductions
were calculated for each of the regulatory alternatives described
in Section 5.2. The control devices (and their efficiencies) at
each emission point considered were respectively: (1) acid-water
scrubber for sterilizer vent (99 percent); (2) recirculating-
fluid vacuum pump for sterilizer vent drain (100 percent);
(3) acid-water scrubber for chamber exhaust (assumed to be
98 percent), and (4) a gas/solid reactor for aeration room vent
(99 percent).
The total nationwide estimated potential emission reductions
and residual emissions for each of the five regulatory
alternatives are presented in Table 6-1.
6.1.2 Secondary Impacts
Secondary air pollutants are those emissions that are not
usually associated with an uncontrolled facility but result from
the use of pollution control equipment (i.e., the control of one
6-1
-------
TABLE 6-1. NATIONWIDE AIR IMPACTS
Regulatory
Alternative
A
B
C
D
E
Nationwide
emission
reduction, %
99
97
94
90
89
EO emission
r educ t ion, Mg / y r
(tons/yr)
1,061 (1,170)
1,042 (1,148)
1,004 (1,107)
963 (1,062)
952 (1,049)
EO residual
emissions, Mg/yr
(tons/yr)
ll! (12)
30 (33)
68 (75)
109 |(120)
120 1(132)
Source: • U. S. EPA Ethylene Oxide Commercial Sterilization Data
Base, 1986, 1988.
6-2
-------
pollutant results in the production of another pollutant;.
Secondary air pollutants are not associated with the use of acid-
water scrubbers, recirculating-fluid vacuum pumps, or gas/solid
reactors.
6.2 WATER QUALITY IMPACTS
If an acid-water scrubber is used to control EO emissions,
there may be water quality impacts, depending on how the spent
scrubber solution (predominately ethylene glycol) is disposed.
Ethylene glycol is generated when the EO exhaust stream contacts
and then reacts with the acid-water solution in the scrubber.
When this solution is spent, the scrubber tank must be emptied
and a fresh acid-water solution added. Each tank initially holds
about 220 gallons (833 liters) of a 10 percent (by volume)
aqueous sulfuric acid (H2SO4) solution, which is neutralized with
50 percent (by weight) caustic (NaOH) before the tank is
drained.3 (See Appendix E Section E.2 [sample calculations for
acid-water scrubbers, assumption 3] for a sample calculation of
the amount of ethylene glycol solution produced per pound of EO
entering the scrubber.) The amount of ethylene glycol solution
produced was calculated based on the assumption that the scrubber
would be drained after 907 kilograms (kg) (2,000 pounds [lb]) of
EO were treated, resulting in a 64 percent (by weight) aqueous
solution of ethylene glycol.
Several methods for the final disposal of the ethylene
glycol were examined. The ethylene glycol produced by the
scrubber can be removed by a waste disposal company, sent to a
municipal wastewater treatment plant, or shipped to a recovery
plant. Removal of the ethylene glycol by a waste disposal
company may not be economically practical for all of the
facilities; this disposal method could account for a high
percentage of the annual operating costs.3 Sending the ethylene
glycol solution to a municipal wastewater treatment plant also
would not be feasible for all sterilization facilities. Some
municipal treatment facilities restrict the concentration level
and amount of ethylene glycol allowed in the discharge to the
wastewater treatment plant.3
6-3
-------
The final disposal method examined was shipping the ethylene
glycol solution to a recovery company. At least three companies
i
accept the ethylene glycol solutions for recovery on a no-credit,
no-cost (except for shipping) basis.4'6 Shipment of the ethylene
I
glycol solution to a recovery company is a disposal method that
would be applicable to all EO users and would not result in any
wastewater impacts. Therefore, the nationwide wastewater impacts
1
calculated for this control device shown in Table 6-2 represent
the maximum potential wastewater impacts. Because of ethylene
glycol recycling, the actual impacts are expected to beilower.
6.3 SOLID WASTE IMPACTS I
Solid waste impacts could occur if the owners or operators
of EO sterilization facilities choose to landfill spent!reactant
from the gas/solid reactor used to control aeration room
emissions. However, the reactant replacement costs were
developed based on spent reactant being returned to the!vendor
for recycling because this alternative was more cost-effective.
Therefore, the nationwide solid waste impacts calculated for this
control device shown in Table 6-2 represent the maximumI potential
solid waste impacts. Because of spent reactant recycling, the
i,
actual impacts are expected to be lower. ;
Additionally, if an owner or operator of an EO sterilization
facility chooses to control emissions with a catalytic oxidizer,
solid waste impacts may occur if the spent catalyst is j
landfilled. j
6.4 ENERGY IMPACTS j
The energy requirements for acid-water scrubbers are
considerably less than those of the gas/solid reactor fan.
Energy requirements for each of the regulatory alternatives are
also presented in Table 6-2. If a catalytic oxidizer is used at
a facility to control emissions, energy impacts will be!
considerably higher.
6-4
-------
tn
ND ENERGY IMPACT!
<:
w"
FH
EWATER, SOLID WAS'
1
Q
M
is
o
H
PH
<
Z
H
W
s
OH
*
CM
1
VO
M
1
EH
I ^ ^^
Total annual
electricity
consumption,
Kw/hr (OOO's
Total annual
solid waste
production,
Mg/(tons)c d
I^J
flj ^| f*^
j3 Q} fe —
C 4J 2
c (C'^j ^
O (C
CD p^ i^
rH 4J 5
(C W g
,; 2 v«n
EH a«e
Q)
« ^
ft -H
•P
rH C
D1^)
<
0
O
VO
•«*•
en
o
CM
O
H
o"
0
O
*!,
0\
VO
in
0
in
CM
<
O
o
•CM
in
H
o
o
o
^
VO
vo
in
o
•*
CM
CQ
O
O
CM
t-H
in
H
O
O
o
^
vo
VO
in
o
»*
CM
O
O
O
0*
0
o
^
VO
vo
in
o
••*
CM
Q
O
O
o"
o
o
^
^
VO
in
o
CM
CM
H
O
o
o en
u
rH C
CJ2C
C tC C
Q) 4J (C
H fl) 4J
43 <4-i (C
•POO)
0) h
•O
faj fll ^tJ
O W O
o
^n fid ^^
CSC
•H O -H
rH O i-H
O O
>i W >i
0-H O
Q) 0)
(It Ij JIJ
O fl) O
0) QJ CO
W 4J W
3 10 3
(0 (C (0
O 5 O
0) 0)
A
U M U
•H CU *H (0 x;
O4 4^ CU 4^
6
-------
6.5 OTHER ENVIRONMENTAL CONCERNS j
6.5.1 Irreversible and Irretrievable Commitment of Resources
i
Regulatory compliance would not preclude the development of
future control options nor would such compliance curtail any
beneficial use of the environment. No long-term environmenta1
losses would result from regulatory compliance by commercial
sterilization facilities.
6.5.2 Environmental Impact of Delayed Standards ,
Delaying the standards would result in possible solid waste
impact reductions, but the reductions would be minimal compared
with the air quality benefits attributable to promulgation of the
standards. There do not appear to be any emerging emission
control technologies that achieve greater emissions reductions or
have significantly lower costs than the control devices '
considered here. Consequently, there are no benefits or
l
advantages to delaying the proposed standards.
6.6 REFERENCES FOR CHAPTER 6 ;
1. Telecon. Soltis, V., MRI, with J. Jorkasky. HIMA.;
Discussion about predicted growth in the ethylene oxide
sterilization industry. March 2, 1987. I
2. Telecon. Hearne, D., MRI, with J. Jorkasky. HIMA.;
Discussion about predicted growth in the ethylene oxide
sterilization industry. May 20, 1991. !
3. Memorandum from Srebro, S., MRI, to D. Markwordt. |EPA:CPB.
Recommendation for costing the disposal of ethylene glycol
produced by the Damas ethylene oxide scrubber. Ju^y 8,
1986. j
i
4. Telecon. Srebro, S., MRI, with J. Hoffman. Med-Chem
Reclamation, Inc. March 6, 1989. Discussion about recovery
of ethylene glycol from ethylene oxide scrubbing liquor.
5. Telecon. Srebro, S., MRI, with K. Dalton. High Valley
Chemicals. March 16, 1989. Discussion about recovery of
ethylene glycol from ethylene oxide scrubbing liquor.
6. Telecon. Srebro, S., MRI, with J. Duvow. Chemstreams, Inc.
March 16, 1989. Discussion about recovery of ethylene.
glycol from ethylene oxide scrubbing liquor.
6-6
-------
7.0 EMISSION CONTROL COSTS
This chapter presents a summary of the methodology used to
develop emission control cost estimates for 188 of the
196 facilities (8 State department of agriculture mobile beehive
fumigation units were not included in this cost analysis) in the
1989 EPA commercial sterilization (CS) data base. These 188 CS
facilities operate a total of 404 sterilization chambers. A
method for estimating EO emission control costs for sterilizer
vents at CS facilities is presented in Section 7.1. Cost
information for chamber exhaust and aeration room controls is
discussed in Sections 7.2 and 7.3, respectively. The results of
the cost analyses are presented in Section 7.4, and Section 7.5
presents other cost considerations.
Costs are further explained in Appendix E. Appendix E
includes: (1) costs for acid-water scrubbers (Section E.I);
(2) sample calculations of the equations used to develop capital
and annual costs for acid-water scrubbers (Section E.2);
(3) aeration room cost analysis (Section E.3); (4) capital and
annual control costs for the sterilizer chamber, chamber exhaust,
and aeration room vent(s) at an example facility (Section E.4);
(5) a breakdown of manifolding costs for these three vents
(Section E.5); and (6) the cost indices and conversion factors
used to convert costs to fourth quarter 1987 dollars
(Section E.6).
Costs presented in this chapter are in fourth quarter 1987
dollars and are for existing facilities only. No new facilities
are anticipated.
7-1
-------
7.1 STERILIZER VENT CONTROL COSTS
This section describes a method for estimating emission
control costs for sterilizer vent(s) and vacuum pump drains at
commercial sterilization facilities. Acid hydrolysis (i.e.,
acid-water scrubbing) was chosen as the basis for the costing
procedure because this control technique currently is known to be
used at 28 commercial sterilization facilities, representing a
variety of industries and a wide range of sterilizer chamber
sizes (7 to 274 cubic meters [m3] [264 to 9,800 cubic f
-------
7.1.1 Description of Components Costed
The following components were costed: (1) an acid-water
scrubber, (2) water-sealed vacuum pump(s) with closed-loop
recirculation, (3) piping for manifolding the chambers to the
existing control device or to a scrubber, (4) operating materials
(i.e., chemicals and chlorine filters), (5) scrubber effluent
disposal, and (6) labor.
Scrubber prices (freight-on-board [F.O.B.]) are listed in
Table E-l. The capital costs associated with the scrubber and
the piping system (for manifolding) are presented in Appendix E,
Sections E.2 and E.5, respectively. The costs of operating
materials, as well as the shipping charges used for computing
disposal costs for the spent scrubber solution, are also
presented in Appendix E, Section E.2.
7.1.2 General Assumptions
Scrubbers were not costed for facilities that had existing
sterilizer vent control devices with efficiencies greater than or
equal to 98-percent. Because of the chemical/physical
characteristics of ethylene oxide as explained in Section 3.4.1,
control devices operating at 98-percent efficiency (e.g., flares)
may be assumed to operate at 99-percent efficiency. If a
facility had a control device with an efficiency below
98 percent, a scrubber to remove 99.0 percent of the remaining
emissions was costed. If a facility had an existing control but
had uncontrolled chambers, it was assumed that the uncontrolled
chambers were manifolded to the control device and that the
existing control device had the capacity to control the
additional emissions from the uncontrolled chambers.
Chamber volume was used as the basis for scrubber sizing.
The relationship of chamber volume to a typical scrubber size is
presented in Table E-l.
If a facility had three or more sterilization chambers, the
scrubber costed was chosen based on the sum of the volumes of the
two largest chambers at that facility. This methodology
simulates the cost of controlling emissions from a facility if
two chambers at that facility were to be evacuated
7-3
-------
simultaneously. If a facility had two chambers, the sc|rubber was
selected based on the volume of the larger chamber. For
facilities with only two chambers, it was assumed that ^he
sterilization cycles could be staggered so that the chambers
would not be evacuated simultaneously.5 \
Costs were developed for recirculating-fluid vacuum pumps to
control drain emissions normally associated with once-through,
water-sealed vacuum pumps. Each sterilization chamber ^t a
facility was costed for a recirculating-fluid vacuum pump unless
the chamber had a control device that utilized a recirculating-
fluid vacuum pump or a recirculating-fluid vacuum pump was
already in place. ;
Piping costs for existing low-efficiency control devices
were calculated based on the assumption that all sterilizers at a
facility could be manifolded to the existing control device. An
acid-water scrubber could then be manifolded to the existing
control device to handle the remaining emissions. For [facilities
that had existing high-efficiency control devices, piping was
costed to manifold any uncontrolled chambers to the existing
control device. Otherwise, piping was costed to manifold all
sterilizer chambers to an acid-water scrubber.
The disposal cost for the aqueous ethylene glycol solution
produced by the acid-water scrubbers was computed as the cost to
ship the solution, either in 55-gallon drums or in a tank truck,
depending on quantity, to a recovery facility. Three recovery
facilities that will accept the ethylene glycol on a no cost/no
payment agreement were identified.6"8 Transportation costs were
calculated by assuming that most CS facilities (except jthose in
Puerto Rico) are located within 1,000 miles of one of the three
known recovery facilities.5 This method of disposal was chosen
because an earlier investigation of alternative disposal methods
indicated that (1) discharging ethylene glycol to municipal
wastewater treatment plants is a disposal method that may not be
available to all sterilization facilities and (2) hauling by a
waste disposal company would be costly for most sterilization
facilities.
7-4
-------
7.2 CHAMBER EXHAUST CONTROL COSTS
This section describes a method for estimating control costs
for chamber exhausts. Although three types of control devices
(i.e., catalytic oxidizers, gas/solid reactors, and acid-water
scrubbers) are theoretically applicable to chamber exhaust
emissions, none are currently in operation controlling chamber
exhaust emissions. The costs of all three control devices were
estimated. However, only the costs of the least expensive
option, acid-water scrubbers, are presented here. Normally, an
efficiency of 99.0 percent is used for an acid-water scrubber
controlling sterilizer vent emissions. However, due to the
differences in emission stream characteristics between the
sterilizer vent and the chamber exhaust, an efficiency of
98 percent was used as a best-case estimate for an acid-water
scrubber controlling the chamber exhaust.
Because both the sterilizer chamber and chamber exhaust vent
use acid-water scrubbers to control EO emissions, the
calculations used to determine annual operating costs (in
Appendix E, Section E.2) for chamber exhaust controls are similar
to those for the sterilizer vent. The main difference is that
the acid-water scrubber for the chamber exhaust controls only
2 percent of the total facility EO use at an efficiency of
98 percent, whereas the sterilizer vent scrubber controls
95 percent of the total facility EO use at an efficiency of
99 percent. Capital costs for the chamber exhaust are based on a
scrubber sized to control a flow rate of either 84 or 168 m3/min
(3,000 or 6,000 ft3/min) (shown in Table E-l), ductwork for
manifolding the chamber exhaust(s) to a common control
(Table E-15), and associated installation costs. A breakdown of
control costs for the chamber exhaust is included in Table E-ll.
7.2.1 Description of Components Costed
The following components were costed: (1) an acid-water
scrubber, (2) ductwork for manifolding vents, (3) operating
materials, (4) scrubber effluent disposal, and (5) labor.
Because of the extreme differences in ethylene oxide
concentrations and the flow rates emitted from the emission
7-5
-------
vents, manifolding was only costed between multiple emissions
sources of the same type (e.g., manifolding multiple sterilizer
vents or chamber exhaust vents or aeration room vents).;
7.2.2 General Assumptions I
In November 1989, the Health Industry Manufacturers
Association (HIMA) conducted a survey of 14 companies !
(23 facilities) to determine the prevalence of chamber exhaust
use. Although these facilities represent only 12 percent of the
facilities in the EPA CS data base, they represent 40 percent of
the annual EO use. Of the 156 chambers these companies operate,
35 chambers (22 percent) do not have chamber exhausts.9\
Typically, these chambers are small in size, i.e., less than 7 m3
(250 ft3). Therefore, based on this HIMA survey, sterilizer
chambers that were smaller than 7 m3 (250 ft3) were not'assumed
to have chamber exhaust and were not costed for control|devices.
I
It was assumed that these facilities can perform more air washes
to reduce worker exposure to EO, which is the purpose of the
chamber exhaust. i
A flow rate of 84 m3/min (3,000 ft3/min) was assumed for
chamber exhaust emissions and served as the basis for sizing an
acid-water scrubber to control chamber exhausts. For this cost
analysis, if a facility has more than two sterilizers, the total
emission flow rate to the control is assumed to be 168 m3/min
_ i
(6,000 ftj/min). This methodology simulates the control cost if
two sterilizers were to vent to the chamber exhaust ;
simultaneously. If a facility has one or two sterilizers, the
emission flow rate is assumed to be 84 m3/min (3,000 ftf/min).
This methodology is based on the assumption that a facility with
only two sterilizers will rarely need to vent the chamber
exhausts simultaneously and is consistent with the methodology
used to develop sterilizer vent costs.5
The scrubber effluent disposal costs for the chamber exhaust
scrubber are the same as those for the sterilizer vent control.
7.3 AERATION ROOM CONTROL COSTS •
This section describes a method for estimating emission
control costs for aeration rooms at CS facilities.
.1
7-6
-------
Initially, costs were developed for gas/solid reactors and
catalytic oxidizers to control aeration room emissions. Because
the gas/solid reactor was the most cost-effective of the controls
considered, it was selected as the basis for the cost estimating
procedure. However, catalytic oxidizers are equally viable and
effective control devices; therefore, cost estimating tables for
catalytic oxidizers, similar to those developed for gas/solid
reactors, are included in Appendix E (Tables E-9 and E-13). For
both of these control devices, facilities and test reports
reported a 99-percent efficiency; therefore, 99 percent was used
as a best-case estimate of control device efficiency.
A gas/solid reactor was costed for 185 of the 188 CS
facilities. Three facilities already controlled aeration room
emissions with catalytic oxidizers and were, therefore, not
included in the cost analysis. A breakdown of the costs of
controlling aeration emissions with gas/solid reactors is
included in Table E-8. A breakdown of the manifolding costs is
shown in Table E-16.
7.3.1 Description of Components Costed
The following components were costed: (1) gas/solid
reactor(s), (2) insulated shipping containers (aeration chambers)
to take the place of aeration rooms, (3) ductwork to manifold the
aeration chambers to a common control device, (4) operating
materials (including utilities), and (5) labor.
7.3.2 General Assumptions
Aeration chambers were costed to replace existing aeration
rooms because emission flow rates from aeration rooms are
typically high (greater than 280 m3/min [10,000 ft3/minj) with
very low (less than 2 parts per million [ppm]) EO concentrations.
By reducing the aeration room size, the emissions can be more
easily controlled because the flow rate is decreased, and the EO
concentration is increased. Therefore, it was assumed that
aeration chambers could be used to replace all existing aeration
processes. The aeration chambers were assumed to be unheated
and, consequently, at ambient temperature. No decrease in
aeration time was attributed to temperature for this cost analysis.
7-7
-------
A regression analysis was performed on aeration room data
received from 35 facilities.10 This analysis was used to
!
correlate the number of model aeration chambers (necessary to
replace existing aeration processes) to a facility's EO use so
that aeration chambers could be assigned to facilities to replace
existing aeration rooms.
The result of the correlation is that facilities that use
2.27 Mg/yr (5,000 Ib/yr) or less of EO were assigned one aeration
I
chamber; however, no cost was attributed to this unit because
facilities that use this amount of EO typically have aeration
rooms smaller than the aeration chamber. Facilities that use
2.27 to 4.54 Mg/yr (5,000 to 10,000 Ib/yr) EO were costed for one
aeration chamber. Facilities that use between 4.54 andi
36.3 Mg/yr (10,000 and 80,000 Ib/yr) EO were costed for!aeration
chambers based on the equation Y = 1.35 x (10~4)X. Facilities
that use more than 36.3 Mg/yr (80,000 Ib/yr) were costed for
aeration chambers based on the equation Y = 2.9 (10~5)Xi + 8.6
(where Y = number of aeration chambers and X = annual EO use,
Ib/yr, for both equations). . ;
Cost estimates for a 28 m3/min (1,000 ft3/min) gas/solid
reactor were obtained from a vendor. This gas/solid rekctor is
capable of handling EO emission concentrations up to lob ppm.11
Costs were extrapolated using an equation provided by the vendor
to determine the cost of an 84 m3/min (3,000 ft3/min) gas/solid
reactor. It was assumed that a maximum of 12 aeration bhambers
(2 parallel sets of 6 in series) could be vented to a cpntrol
unit. The emission flow rate from each of these aeratijon
chambers was assumed to be 7 m3/min (250 ft3/min), and the number
of control units was selected so as to minimize capital; costs.11
Disposal costs for the spent gas/solid reactant were
calculated for a transportation distance of 1,500 miles. This
cost was based on the assumption that the reactant could be
returned to the vendor for recycling on a no-charge, nofcredit
basis. A distance of 1,500 miles was used because most!
facilities (except those in Puerto Rico) are located wijthin
1,500 miles of the manufacturer of the reactant.11 ',
7-8 i
-------
7.4 RESULTS OP COST ANALYSIS
The nationwide cost impacts associated with each of the
regulatory alternatives are shown in Table 7-1. Table 7-2, shows
the cost impacts on three representative facilities. A
"facility" in this case includes the sterilization vent, chamber
exhaust vent, and aeration room vent emissions. These facilities
were selected to represent the median facility (with regard to
annual EO use and cumulative sterilizer chamber volume) in each
of the following annual EO use ranges: <272, 272 to 18,150, and
>18,150 kg/yr (<600, 600 to 40,000, and >40,000 Ib/yr). A
representative facility was not selected for facilities that use
less than 11 kg/yr (25 Ib/yr) EO because such a facility does not
exist. The facilities that were selected represent small,
medium, and large facilities that use 228, 3,963, and
67,604 kg/yr (504, 8,736, and 149,000 Ib/yr) of EO; and have
cumulative sterilizer volumes of 5.7, 28, and 112 m3 (204, 1,000,
and 4,002 ft3), respectively.
7.5 OTHER COST CONSIDERATIONS
In addition to the costs described above, costs to comply
with other Federal rules or regulations may be incurred by
commercial sterilization facilities. These costs are described
in the following section.
At CS facilities where workers handle or are near product
during the sterilization process, measures must be taken to
reduce EO worker exposure to less than 1 ppm per 8-hour time-
weighted average concentration. In most facilities with
cumulative sterilizer volumes less than 7 m3 (250 ft3) worker
exposure is minimized using the chamber exhaust, which evacuates
EO-laden air from the chamber while workers are loading/unloading
the sterilizer chamber. Two of the regulatory alternatives
reguire add-on controls for the chamber exhaust. If a facility
were to elect to disable the chamber exhaust in lieu of the add-
on control there would be additional costs to ensure continued
OSHA compliance.
7-9
-------
H
W
o
o
CO
60
•i §
W o «.
"8
U
It
•— '-^
o -3
3 g
II
8
S
o o
7-10
-------
•> O oo
Sgg*
a -5 »
SJ1
2CJ.
S
10
J*
It
± §
JBl
51
CO
EH
CO
O
U
EH
H
J
H
H
I
CO
0
Z 2 °
< •< <
< •< <
z zz
zzz
333
<; oo oc
z zz
zzz
zzz
1
r>a
t^
a?2
z 9 -
z 3 -
I
I
z •"
fs
z -
z ••
oe
•^
I
u.
U
7-11
-------
7.6 REFERENCES FOR CHAPTER 7
|
1. Memorandum. deOlloqui, V., to Commercial Sterilization
files. Responses to the July 1986 Section 114 Letter.
August 22, 1990. List of facilities that responded to the
July 1986 Section 114 letter regarding the use of ethylene
oxide by miscellaneous sterilization and fumigation
facilities. |
i
2. Memorandum. deOlloqui, V., to Commercial Sterilization
files. Responses to the July 1988 Section 114 Letter.
August 22, 1990. List of facilities that responded to the
July 1989 Section 114 letter to 39 facilities with maximum
individual risks greater than 10". [
i
3. Memorandum. Srebro, S., to D. Markwordt. EPA/CPB;
Baseline Cost of Reducing Ethylene Oxide Emissions from
Sterilizer Vents and Associated Vacuum Pump Drains;
November 15, 1990. i
i
4. Memorandum. Srebro, S., to D. Markwordt. EPA/CPBi.
Examination of Ethylene Oxide Control Efficiencies!
October 5, 1988. ;
5. Memorandum. Srebro, S., to D. Markwordt. EPA/CPBl Cost
Effectiveness of Reducing Ethylene Oxide Emissions from
Sterilizer Vents and Associated Vacuum Pump DrainsI
March 21, 1991. j
I
6. Telecon. Srebro, S., MRI, with J. Hoffman. Med-Chem
Reclamation, Inc. March 6, 1989. Discussion about recovery
of ethylene glycol from ethylene oxide scrubbing liquor.
7. Telecon. Srebro, S., MRI, with J. Duvow. Chemstrkams, Inc.
March 16, 1989. Discussion about recovery of ethylene
glycol from ethylene oxide scrubbing liquor. j
i
8. Telecon. Srebro, S., MRI, with K. Dalton. High vklley
Chemicals. March 16, 1989. Discussion about recovery of
ethylene glycol from ethylene oxide scrubbing liqubr.
9. Letter and confidential attachments from Jorkasky JT., Health
Industry Manufacturers' Association, to S. Srebro,! MRI.
December 1989. Information on use of rear chamber} exhaust
at medical equipment suppliers. \
10. Memorandum. deOlloqui, V., to Commercial Sterilization
files. Responses to the July 1989 Section 114 Letter.
August 22, 1990. List of facilities that responded to the
July 1988 Section 114 letter to 44 facilities regarding
sterilizer operating parameters, vacuum pumps, control
devices, and aeration room data. i
7-12
-------
11. Memorandum. deOlloqui, V., and S. Srebro. MRI, to
D. Markwordt. EPA/CPD. March 21, 1991. Costing
Methodology for the Control of Aeration Room Emissions.
7-13
-------
-------
8.0 THE ECONOMIC IMPACTS OF THE CANDIDATE NESHAP CONTROLS
8.1 INTRODUCTION
Companies performing ethylene oxide (EO) sterilization fall
into two general groups: in-house sterilizers and contract
sterilizers. In-house sterilizers are companies that produce the
goods needing sterilization. As part of their production
process, they also sterilize the products. Also included in this
group are laboratories, museums, and libraries. These in-house
sterilizers do not produce goods needing sterilization, but, like
the other in-house sterilizers, sterilization is a small but
necessary part of their operations. Museums, for example,
specialize in preserving and displaying artworks or artifacts.
To be preserved, some of these artifacts must be sterilized or
fumigated. Sterilization is only a very small part of the
activities carried on by museums; some of them choose to perform
it onsite, while others send their artifacts offsite to a
contract sterilizer.
Contract sterilizers are companies that specialize in
sterilization/fumigation, so sterilization is a major part of
their business. They do not, in general, produce any of the
goods being sterilized; rather, they offer the service of
sterilization to other producers.
This chapter identifies the industries affected by this
regulation and evaluates the economic impacts of three possible
control options. First, the industries that perform EO
sterilization are described; then the industries producing
8-1
-------
products requiring sterilization are profiled. Finally, the
effects of the control options on these industries are assessed.
8.2 ETHYLENE OXIDE STERILIZATION
8.2.1 Process Inputs
The major capital equipment requirement for sterilization is
the sterilization chamber. The 188 facilities covered
in this
profile operate 404 chambers, an average of 2.2 chambers per
facility. The number of chambers per facility ranges from one to
nine. Chamber sizes range from 0.01 cubic meters (m3)
(0.35 cubic feet [ft3]) to 177 m3 (6,250 ft3) and average about
16 m3 (565 ft3).1 A typical chamber has a useful life of
approximately 10 years; at the end of that time, the salvage
value of the chamber is less than 1 percent of the initial
capital investment.2 ;
Certain design characteristics of the sterilization chamber
are determined by the gas to be used for sterilization.; These
design characteristics limit the possibility of switching gases
in the short run. When pure EO is used, the chamber :
instrumentation and room must be explosion-proof.3 Pure EO
chambers must also operate with a deep vacuum to rid the chamber
of oxygen. Chambers that use an EO/CO2 mixture, which ^requires a
substantially higher operating pressure than other gases, have to
meet higher standards of construction to withstand the pressure.4
The various chamber specifications are not mutually
exclusive. For example, a chamber designed to withstand the high
operating pressure associated with the EO/CO0 mixture can also
\
accommodate the lower pressures required by other sterilant
gases. Therefore, some facilities can have chambers with the
characteristics necessary for use with more than one type of
sterilant gas. Consequently, these facilities can alternate
between gases to achieve optimal combinations of product and
sterilant gas for each sterilization cycle. Facilities! involved
in testing and research and those that sterilize a wide variety
of products are more likely to operate chambers in this manner.
Other equipment required for the sterilization pro;cess
includes: ;
8-2
-------
1. A pump to create a vacuum in the chamber; and
2. A pump to force sterilant gas or air into the chamber.
Some facilities also have a separate room for the aeration
step of the process, although aeration at some facilities can be
accomplished in the sterilization chamber so that products do not
have to be moved.
Labor requirements for EO sterilization are usually higher
than for other types of sterilization (e.g., thermal or
radiation). Each sterilization cycle must be closely monitored
because several critical process variables require careful
attention.5 In addition, sterilization with pure EO requires
strict safety precautions and extensive monitoring by facility
personnel because pure EO is flammable.
8.3 SUBSTITUTION POSSIBILITIES AND THE PRICE ELASTICITY OF
DEMAND
The extent of substitution between the EO-based gases is
limited by the characteristics of the sterilization chamber and
the compatibility of the sterilizing medium with the products
being sterilized. As noted earlier, both pure EO and the EO/CO,
mixtures may require specially designed sterilization chambers.
Chambers may be modified to use pure EO, but industry standards
prohibit modifying chambers for use with the EO/CO2 mixture.
However, the different pressures under which sterilization is
performed for each sterilant gas can damage some products or
packaging. Consequently, substituting with sterilant gases that
require extremely high or low operating pressures is limited by
the characteristics of the products being sterilized.
Gamma radiation sterilization can substitute for EO
sterilization for many products. Unlike EO, radiation can
sterilize liquids and products in vapor-tight packages; however,
it discolors plastics and damages Teflon and acetyl delrin.4'5
Gamma radiation cannot be used to sterilize Pharmaceuticals
because the radiation may alter the chemical structure of the
drugs. Gamma radiation is expected to make some additional
inroads into the EO market for sterilization.2 Although gamma
radiation is a likely substitute for EO sterilization, safety
8-3
-------
concerns regarding transportation and disposal of the radiation
source and the cost of the radioactive cobalt used as a radiation
source will probably limit any increase in the percentage of the
products sterilized by radiation. [Several other chemical
substitutes for EO exist, but their use is more limited than
gamma radiation; the substitution possibilities for these
chemicals are discussed in later sections in conjunction with the
appropriate end products.] i
The importance of the sterilization procedure in the overall
production process of many products presumably has a profound
effect on its demand elasticity (i.e., the responsiveness of
quantity demanded to a change in the price of sterilization
services). Four main issues influence the elasticity of demand
for a factor within an industry.6 Specifically, the elasticity
of demand should vary directly with the elasticity of demand for
the final product, the factor's share of the costs of production,
the elasticity of supply of other factors, and the elasticity of
substitution between the factors. j
These influences suggest that the demand for sterilization
within the relevant industries is relatively inelastic.; This
conclusion is based in part on the low ratio between this cost of
sterilization and the total cost of production.7 Additionally,
the elasticity of substitution between factors of production
i
within the industries is relatively low; precautions taken to
minimize contamination during production do not necessarily
i
lessen the need for sterilization but enhance the effectiveness
of the process. Generalizations regarding the elasticity of
demand for the final product and the elasticity of supply of
other factors are not possible due to variations from industry to
industry.
8.4 SUPPLY OF EO STERILIZATION SERVICES !
This section profiles the facilities that sterilize/
fumigate medical devices and other miscellaneous produces with
EO, excluding sterilization activities in hospitals. For
simplicity, we refer to this process simply as sterilization.
For a variety of reasons discussed later in this section,
8-4
-------
sterilization does not form a cohesive industry but is instead a
part of the production process in several industries. Therefore,
this profile will be limited in the scope of its analysis. The
most important limiting factor is the lack of data on
sterilization as a separate step in the production process.
Consequently, commercial sterilizers are grouped here by
industry; sterilization is then analyzed within the context of
that industry.
8*4*1 National Summary of Ethvlene Oxide Sterilization
The economic analysis presented in this chapter covers
188 sterilization facilities for which data are available. About
half of the 188 facilities are suppliers of medical devices or
other health-related items; the other facilities engage in
several miscellaneous sterilization and fumigation operations
(discussed later).
In 1988, approximately 1,913 megagrams (Mg) (4.22 million
pounds [lb]) of EO was used for sterilization purposes by the 188
facilities covered in this analysis. [The EPA commercial
sterilization database contains 1988 data for 32 facilities, the
remainder are 1985 or 1986 values; for simplicity, 1988 will be
referred to as the base year.] Table 8-1 presents some summary
statistics on the use of the sterilant gases at the
188 facilities, separated into two categories—EO use and total
gas use. The tremendously wide range in the use of sterilant gas
per facility is noteworthy.
As mentioned above, several characteristics of the
sterilization process make it difficult to profile the process as
a discrete industry. The most important characteristic stems
from the role of sterilization in the overall production process.
Except for contract sterilizers, the sterilization process is an
intermediate step in the production process. Therefore,
separating sterilization from the production of the sterilized
products is difficult. Another difficulty is the absence of a
Standard Industrial Classification (SIC) code listing for
sterilization either separately or within the classifications for
the industries that employ this process. Furthermore, EO
8-5
-------
TABLE 8-1. SUMMARY STATISTICS ON THE USE OF STERILANT GAS
AT 188 COMMERCIAL STERILIZATION FACILITIES, 1988
Total use
Use per facility
Average
Range
Use per chamber
Average
Range
Mg/yr (tons/yr)
kg/yr (Ib/yr)
kg/yr (lb/yr)
kg/yr Gb/yr)
kg/yr (lb/yr)
Ethylene oxide
1,913 (1,883)
10,179 (22,441)
1-129,090 (2-284,594)
4,737 (10,443)
0.1-62,045 (0.2-136,786)
Total gas
! 6,560 (6,456)
34,897 (76,935)
6-359,400 (13-792,341)
16,239 (35,801)
0.1-168,000(0.2-370,382)
8-6
-------
sterilization accounts for only 50 to 60 percent of all
sterilization activities.4 Finally, the diversity and
specialization of the industries that sterilize products limit
the amount of data available for this profile.
8.4.2. Industry Groups Supplying EO Sterilization Services
Several main categories of facilities sterilize some portion
of their output:
1. Medical device suppliers;
2. Other health-related suppliers;
3. Pharmaceutical manufacturers and other drug-related
manufacturers;
4. Spice manufacturers and other food-related
manufacturers;
5. Museums and libraries;
6. Laboratories (research, testing, and animal breeding);
and
7. Contract sterilizers.
Table 8-2 summarizes the specific SIC codes associated with
these industry categories and the number of facilities in each.
As might be expected from the large number of SIC codes, the
188 sterilization facilities sterilize a wide variety of
products. These products include surgical gloves and hypodermic
needles sterilized by medical device suppliers, books fumigated
by libraries and museums, and spices fumigated by spice
manufacturers.
The sterilization processes used by the above industries
have several similarities. However, a distinction can be made
between in-house sterilization and contract sterilization. A
majority of the facilities covered in this profile operate a
sterilization chamber at the same location as the remainder of
the production process. The exception to this rule is the subset
of commercial sterilizers that sterilize products for other
companies on a contract basis. Not only do contract sterilizers
accept a variety of products for sterilization, but they may also
supervise the final distribution of the products. It should be
noted that the distinction between these two types of
8-7
-------
TABLE 8-2. STANDARD
188 COMMERCIAL
INDUSTRIAL CLASSIFICATION CODES FOR
STERILIZATION FACILITIES, 1988
SIC Code
Medical device suppliers « 62
3841
3842
Other health-related supplier! « 24
3693
5086
2211
2821
2879
3069
3569
3677
3999
Pharmaceutical manufacturer* = 39
2834
5122
2831
2833
Spice manufacturers — 23
2099
5149
2034
2035
2046
Museums and libraries = 13
8411
8231
Laboratories = 10
2790
7391
8071
8922
7397
Contract sterilizers — 17
7399
7218
8091
No. of
facilities
44
18
7
5
4
1
2
1
1
1
1
1
34
2
2
1
17
3
1
1
1
11
2
4
2
1
2
1
14
1
2
Description of category
[
Surgical and medical instruments and apparatus
Orthopedic, prosthetic, and surgical appliances and supplies i
Miscellaneous plastic products ;
Radiographic X-ray, fluoroscopie X-ray, therapeutic X-ray, and other X-ray
apparatus and tubes; electrochemical and electrotherapeutic apparatus
Professional equipment and supplies
Broad woven fabric mills, cotton ' i
Plastics materials, synthetic resins, and nonvulcanizable elastomers
Pesticides and agricultural chemicals, NEC
General industrial machinery and equipment NEC
Electronic coils, transformers, and other inductors
Electronic coils, transformers and other inductors
Manufacturing industries, NEC
Pharmaceutical preparations
Drugs, drug proprietaries, and druggist's sundries
Biological products :
Medicinal chemicals and botanical products <
Food preparations, NEC I
Groceries and related products, NEC
Dried and dehydrated fruits, vegetables, and soup mixes
Pickled fruits and vegetables, vegetable sauces and seasonings, and salad dressings
Wet corn milling
Museums and an galleries
-ibraries and information centers
Animal specialties, NEC
Research and development labs
rfedical labs !
Noncommercial educational, scientific, and research organizations
Commercial testing labs
iusiness services, NEC
ndustrial launderers
iealth and allied services, NEC
NEC - Not elsewhere classified.
8-8
-------
sterilization is not always well defined. Some facilities,
especially those within the medical device suppliers and
pharmaceutical industries, sterilize their own products in-house
and also accept products on a contract basis from other firms.8
8.4.2.1 Medical Device Suppliers. Sterilizers of medical
devices (SIC 3841 and 3842) represent the largest single segment
of commercial sterilizers covered in this analysis, including
62 of the 188 facilities. The total annual output of medical
devices sterilized in the United States is estimated as 15 to
20 billion products, with at least 50 percent of these products
sterilized with EO.1
Some medical devices must be sterilized to be marketed.
Ethylene oxide, especially the 12/88 mixture, is used for medical
device sterilization because of its wide range of effectiveness*
The FDA has set strict guidelines for medical device sterilizers
to ensure that necessary levels of sterility are achieved. These
guidelines, called Good Manufacturing Practices (GMP), include
requirements for such things as preliminary testing, procedural
supervision, quality assurance, and final labeling.5
The U.S. medical device supply industry has been changing
rapidly because of cost containment measures, increased
competition, and changes in the health care system. A trend
toward consolidation among both buyers and sellers of medical
devices has been evident. Many hospitals have formed buyers'
groups or corporate buying arrangements for purchasing supplies.
Consolidation has allowed suppliers to increase their efficiency
and broaden their product and distribution bases. Declining
hospital occupancy and shorter visits have decreased the demand
for medical devices from hospitals. At the same time, demand for
medical devices from outpatient facilities has experienced strong
growth. The number of surgical procedures being performed has
been growing, but a larger share of these procedures is being
done on an outpatient basis.
Product quality is becoming a major issue in the medical
device industry. Under pressure from the FDA and increasing
competition, suppliers are striving to improve their
8-9
-------
manufacturing processes and products. The surgical and medical
instruments industry (SIC 3841) is expected to grow about 7
percent per year (in 1982 dollars) between 1989 and 1993. The
surgical appliances and supplies industry (SIC 3842) is projected
to grow at an annual rate of 8.5 percent during this time.
According to the Census of Manufactures, there were
2,600 establishments in the surgical and medical instruments and
surgical appliances and supplies industries (SIC 3841 and 3842)
in 1987. Table 8-3 summarizes various statistics for these
industries. Product data were collected for all products
classified in either SIC 3841 or 3842 that are produced by all
industries; industry data represent all facilities classified in
either SIC 3841 or 3842, including their output of nonmedical
products.9'10
The data presented in Table 8-3 reveal several trends within
the industry. The total value of shipments for medical device
suppliers is given in current dollars and in 1982 dollars.
Throughout the 1980's, the industry has shown strong, steady
growth. Total employment has risen steadily since 1972,, while
the proportion of production workers has fallen slightly. In
t
1988, production workers made up 63 percent of the total work
force, as compared with 69 percent in 1972. Growth in foreign
markets, along with a lower value of the U.S. dollar relative to
other currencies, has allowed U.S. manufacturers to inqrease
exports of medical devices. Manufacturers of surgical iand
medical instruments (SIC 3841) increased exports by 17 percent
between 1987 and 1988. Imports of these products also increased
but at a slower rate (13.8 percent), allowing the trade surplus
i
to grow by 29 percent. Exports of surgical appliances jand
supplies (SIC 3842) grew 25 percent during this time, while
imports grew only 5.7 percent. Exports of medical devices are
expected to remain high if the U.S. dollar does not appreciate
substantially. Although the U.S. has consistently had an overall
trade surplus in the medical device industry, it continues to
have trade deficits with West Germany and Japan.9
8-10
-------
^
H
O
£
S?
l|
EH ?°
CO
rjo
O H
P^ —41
|S
sl
ii
o S
(i
. RECENT
DEVICE S
n
I
00
TABLE
\
"
00
oo
£
r-
§
2
«•*
oo
*"*
.1
S
~"
I
j
z
1
^
m
oo
1
^
00
••T
*"*
00
g
m
S
£
^^
Value of shipments
g
-•
2
S
»rt
rj
|
^*
£
o"
o\
g
""
s
A
00
^,
|| Value of shipments
»f
Industry data1
Z
r-
*
i
i
t-;
ro
*>•
i
5
0.
00
t-
,« ,
•1
•^
Total employment
Z
r-
8!
s
oo
00
^
00
oo
00
r^
^
f>
5
^
J
^
e
|
o
Z
Z
>r>
^
*
"^
W^
*r>
O
oo
R
~
i
*0
O1
» Capital expenditures
2
*
Ot
>o
o
w
oo"
M
S
**^
OO
00
en
1
"S,
O^
s-ll 1
111 1
£.5° 0.
fi g i
i== 5s
o g M ^ « •
> H -^ e _.
2 g o _ ^ e
J J S S § J
f 5 e " 5 |
"s *
-------
Although a wide range of products.is included under the
broad category of medical devices, generalizing about the
importance of the products is still possible. Medical Equipment
is a basic part of health care service, which is a necessary
service. Therefore, the elasticity of demand for medical devices
should be highly inelastic. However, the trends mentioned above
(increased price competition among health care facilities and
increased use of alternatives to hospitals) may indicate that the
demand for medical devices is becoming more elastic. The
availability of imports may also increase the elasticity of
demand for domestically produced medical devices. |
Table 8-4 shows some summary statistics on the sterilization
chambers and sterilant gases used by the 62 medical device
suppliers included in this study. These 62 facilities pperated a
total of 145 EO sterilization chambers in 1988, an average of
2.3 per facility. The number of chambers per facility varied
from one to eight. Average chamber volume per facility! was
40.1 m3 (1,416.0 ft3) but covered a wide range from 0.03 m3
(1.06 ft3) to 232 m3 (8,193 ft3). In 1988, these 62 facilities
used 665 Mg (654 tons) of EO, slightly under 11 Mg (10 tons) per
facility. Like chamber volume per facility, EO use varied widely
from 0.01 Mg (0.01 tons) to 109 Mg (107 tons). Total gas use,
which averaged 42 Mg (41 tons) per facility, also covered a wide
range; the smallest user reported 0.05 Mg (0.06 tons) and the
largest reported 511 Mg (503 tons).1 !
By subtracting EO use from the total gas use, and then
dividing by EO use, an inert-gas ratio is obtained. This ratio
indicates the extent of reliance on pure EO versus mixeo:
sterilant gases. In particular, a ratio near zero would suggest
extensive reliance on pure EO (because EO use and totaljgas use
would be almost identical), but a much higher ratio would
indicate greater use of sterilant mixtures, such as 12/88.
During 1988, the inert-gas ratio for the 62 medical device
suppliers was 2.88.
8.4.2.2 Other Health-Related Suppliers. Twenty-four
facilities were included in this study that produce some type of
8-12
-------
TABLE 8-4. SUMMARY STATISTICS ON STERILIZATION CHAMBERS AND
GASES USED BY 62 MEDICAL DEVICE SUPPLIERS, 1988
Sterilization chambers = 145
Number per facility
Chamber volume per facility, m^ & '
Ethylene oxide use = 665.6, Mg/yr (655. 1 tons)
Use per facility, Mg/yr (tons)
Total gas use = 2,576.7 Mg/yr (2,536.0 tons)
Use per facility, Mg/yr (tons)
Average
2.3
40.1
(1,416.1)
10.7
(10.5)
41.6
(40.9)
Standard
deviation
1.6
37.9
(1,338.4)
6.9
(6.8)
76.2
(75.0)
Range
1-7
0.03 - 232
(1.06 - 8,193)
<0.05 - 109.1
(<0.05- 107.3)
0.05-511.2
(0.5 - 503)
8-13
-------
health-related supplies but are classified under a morej general
SIC code (see Table 8-2). For example, five facilities! have a
primary SIC code that involves the manufacturing of X-r^y
equipment and nine facilities manufacture various plastic
products. The large number of SIC codes illustrates the
diversity of industries that sterilize health-related equipment
with EO. For this reason, the value of shipments in Table 8-3
was presented on a product basis rather than on an industry
basis. No other data specific to the relevant SIC codes will be
presented. '
Table 8-5 provides some summary statistics on the
sterilization chambers and sterilant gas used by the 24!other
health-related suppliers included in this study. These;
24 facilities operated a total of 53 sterilization chambers in
1988, an average of 2.2 per facility. This is slightly;lower
i
than the medical device suppliers group, which averaged'
2.3 chambers per facility. !
The other health-related suppliers used 276 Mg (271 tons) of
EO in 1988. They averaged 11.5 Mg (11.3 tons) of EO per
facility, covering a range of 0.002 Mg (0.0022 tons) to!129.0 Mg
(127.0 tons). Overall, these facilities used less EO than did
the medical device suppliers. However, their EO use per facility
and their total gas use per facility are slightly higher than
i
those of medical device suppliers. The inert gas ratio|for other
health-related suppliers was 2.46. I
8.4.2.3 Pharmaceutical Manufacturers. The broad category
of pharmaceutical manufacturers includes those facilities whose
SIC code was 2831, 2832, 2834, or 5122. These facilities are all
connected with pharmaceutical preparations or other medicinal or
biological products as manufacturers or, in some cases,!as
wholesalers. However, a vast majority of the facilities are
classified as SIC 2834. Therefore, the profile of pharmaceutical
manufacturers is focused on that industry group. Sterilization
has a variety of uses in this industry, but it is as closely tied
to the ultimate safety and effectiveness of the products as is
the medical device industry. i
8-14 j
-------
8-5. SUMMARY STATISTICS ON STERILIZATION CHAMBERS AND
GASES USED BY 24 OTHER HEALTH-RELATED SUPPLIERS, 1988
Sterilization chambers = 53
Number per facility
Chamber volume per facility, m^ (ft^)
Ethylene oxide use = 275.5 Mg/yr
(271.2 tons)
Use per facility, Mg/yr (tons)
Total gas use = 952.5 Mg/yr, (937.5 tons)
Use per facility, Mg/yr (tons))
Average
2.2
41.3 (1,458.5)
11.5(11.3)
39.7(39.1)
Standard
deviation
1.5
58.4 (2,062.4)
7.07 (6.%)
51.1 (50.3)
Range
1 -6
0.4 - 207.2
(14.1 - 7,317.2)
< 0.05 -129.1
(<0.05- 127.1)
0.05 - 152.2
(0.05-149.8)
8-15
-------
The pharmaceutical industry shows several interesting
trends. Pharmaceutical companies continue to spend increasing
amounts on research and development: they spent $5.4 billion in
1987 and a record $6 billion in 1988. Experts expect that the
high cost of research and development will lead to consolidation,
especially among small specialty manufacturers. Patent piracy
remains a problem for this industry. Manufacturers lost an
estimated $2 billion to patent pirates in 1988. Sales of generic
drugs will continue to grow as patents on existing drugs expire.
Generic drugs currently make up about 12 percent of the
prescription market and will probably account for 30 percent by
1993. The pharmaceutical industry also expects strong growth in
the market for over-the-counter drugs as more drugs are made
available for purchase without a prescription.9
The value of pharmaceutical product shipments rose, an
estimated 2.7 percent (in constant dollars) between 1987 and
1988. The U.S. Department of Commerce predicts that the industry
will grow between 2 and 3 percent a year between 1989 and 1993.
Factors contributing to this growth will include an increasing
demand for drugs by an aging population, greater exports to
developing countries, and improved productivity through;
computerization.9 l
According to the Census Bureau, a total of 718 facilities
were classified under SIC code 2834 in 1987. Only 34 have been
identified as commercial sterilizers.10 The Pharmaceuticals
i
group, as defined in this chapter, includes five additional
facilities from SIC codes 5122, 2831, and 2833. Table J9-6
reports recent performance and forecast data for all ;
pharmaceutical manufacturers classified under SIC code ;2834. The
data are presented in much the same manner as in Table '8-3, with
a distinction between industry and product data. I
As shown in Table 8-6, the value of pharmaceutical shipments
has increased steadily through the 1980's in both nominal and
real terms. Industry employment, after falling in the (early
1980's, has shown an upward trend since 1985. Capital !
expenditures decreased somewhat between 1985 and 1986 but
8-16
-------
W
w
£t?
£4
1
j
8
w
F .
g
W
y
1
0.
£
,4;
SH
Q --*.
•<*
1"H CO
CO CO
«*
Wr \
w
CJ H
O CO
Q
W
O
1
e&
o
fa
w
O4
E-i
g
y
w
C£
•
VO
1
CO
H
1
1
00
oo
Ov
VO
OO
Ov
£
O\
g
o\
!P
r-
o\
1
s
g
ts^
S
00
ts
(N
O
(+}
~
S
*.
oo
Ov
^fc
VO
Sf
s^^-
42
f
•s
^a
1
8
*c\
r^
*•*
a
**
\O
s
oo"
en
oo
t—
-.
S
vT
-
S
2
S
^f
2
yf
3
2
v^
* — '
Value of shipments i
1
i
1
0
B
q
oo
o
a-
o
g
q
(N
q
m
q
en
C>1
q
r4
21
V
tn
o
^
1
I
w
VO
^
S*
1
^
oo'
^
06
^
oo
VO
S
q
T)
^
f
&
pn
O
N«^
1 Production workers
Z
^
I
«M
00
t^
s
CO
-.
q
vo
q
-------
increased in 1987. In recent years, both imports and exports of
Pharmaceuticals have been increasing. i
Table 8-7 contains some summary statistics on the •
sterilization chambers and sterilant gas used by the 39, pharma-
ceutical manufacturers included in this study.1 These j
39 facilities operated a total of 82 sterilization chambers in
1988 , averaging 2.1 per facility. The average chamber Volume per
facility for pharmaceutical manufacturers is 29.2 m3 |
(1,031.2 ft3), with chamber volume ranging from 0.1 m3 ^:o 147 m3
(3.5 ft3 to 5,191 ft3. Although the number of chambers! per
facility in the Pharmaceuticals group roughly equals the number
for the medical device suppliers group, the average chamber
volume per facility for pharmaceutical manufacturers isj about
70 percent of the average for medical device suppliers. j
The pharmaceutical manufacturers used 416 Mg (409 tons) of
EO in 1986, averaging 10.7 Mg (10.5 tons) per facility and
varying from 0.01 Mg (0.01 tons) per facility to 84 Mg !
(82.7 tons) per facility. They also used 830 Mg (818. 9j tons) of
total gas, with an average of 21.3 Mg (21.0 tons) per facility
and a range of 0.06 Mg to 92 Mg (0.06 to 90.5 tons) per| facility.
The ranges of both EO use and total gas use are much narrower for
pharmaceutical manufacturers than they are for medical device
suppliers. Finally, the inert-gas ratio for Pharmaceuticals is
I
1.00, compared to a 2.88 ratio for the medical device suppliers.
8.4.2.4 Spice Manufacturers. Twenty-three of the!
188 facilities in this study have been categorized as spice
I
manufacturers. These firms fall into five SIC codes: 2099,
5149, 2034, 2035, and 2046. However, as shown in Table! 8-2,
about two-thirds of these firms are in SIC code 2099. |
Consequently, the discussion below focuses on this industry
group . " !
About one-quarter of all spices manufactured in the United
States are treated with EO to control fungi, molds, bacteria, and
insect eggs.14 Ethylene oxide pasteurization increases shelf
life and decreases health risks.14 An industry magazinfe reports
that "since the spice processor cannot preclude or remojve
8-18
-------
TABLE 8-7. SUMMARY STATISTICS ON STERILIZATION CHAMBERS AND
GASES USED BY 39 PHARMACEUTICAL MANUFACTURERS, 1988
Sterilization chambers = 82
Number per facility
Chamber volume per facility, m3 (ft3)
Ethylene oxide use = 416.1 Mg/yr (409.6 tons)
Use per facility, Mg/yr (tons)
Total gas use = 830.4 Mg/yr (817.3 tons)
Use per facility, Mg/yr (tons)
Average
2.1
29.2 (1,031.2)
10.7
(10.5)
21.3
(21.0)
Standard
deviation
1.4
20.0 (706.3)
3.12
(3.07)
28.1
(27.7)
Range
1-6
0.1 - 147.3
(3.5-5,201.9)
< 0.05 -129.1
(< 0.05-127.1)
0.06 - 92.2
(0.06 - 90.7)
8-19
-------
microorganisms from spices, the only means to control
microorganisms are: (1) to maintain low moisture content to
hinder growth and activity and (2) to treat spices with EO.'«15
Several substitutes within the spice industry are available
for EO, each of which has disadvantages. Heat sterilization is
useful on only a small number of spices, because it lightens or
darkens spice color and can cause a 15-percent loss in spice
strength by volatizing spice oils. Neither ionizing radiation
nor ethylene imine are approved by the FDA, and their j
I
effectiveness is not known. The use of propylene oxideiis
restricted by the FDA to starches, gums, processed spices, cocoa,
and processed nut meats.14 However, propylene oxide must be
heated to be as effective as EO. Using heat restricts its use to
a small number of spices because heat volatizes oils in!spices,
affecting their guality. Without heat, propylene oxide!has one-
tenth the microbial killing activity of EO. Finally, propylene
I
oxide requires 16 to 48 hours of exposure time, compared to 6 to
8 hours for EO. |
According to an industry source, radiation treatment is a
promising alternative to EO fumigation, having several advantages
over EO.16 First, although EO kills a high percentage of
bacteria, radiation kills all bacteria. Second, EO requires high
humidity and a vacuum, while radiation treatment can be done
under ambient conditions. Third, EO leaves a residue iii all
foods that have been treated, but radiation leaves no residues of
any kind. !
According to the Census Bureau, SIC 2099 had 1,976jfirms in
11 i
1982.x/ As shown in Table 8-8, the value of shipments for SIC
2099 increased steadily from $3.6 billion in 1972 to
$11.0 billion in 1982, a nominal growth of 201 percent over the
10-year period. Value added also increased steadily over the
same period rising from $1.8 billion in 1972 to $5.7 billion in
1982. Employment experienced two low periods, one in the mid-
1970's and another in the early 1980's. The number of production
workers declined in 1975 and again in 1979 and 1980, while total
employment declined in 1979 and 1981. Both employment measures
8-20
-------
a\
a\
o
OJ
C
V.
*•*
tf.
tt
e
c
w
o
w
g
w
u
I
w
u
w
00
I
CO
H
I
S
S3
i
S
8-21
-------
increased from 1981 to 1982. Capital expenditures increased from
$9"l.8 million in 1972 to $295.4 million in 1982 with declining
years in 1977 and 1981. j
In 1987, the industry definition of SIC 2099 changed.
Although spice manufacturers are still included under the new
definition, industry figures for 1987 cannot be compared with
previous years. Table 8-9 presents the 1987 data for SIC 2099.
The spice and seasoning industry relies heavily on foreign
trade. According to one estimate, two-thirds of the spices
consumed in the United States are imported.18 This estimate
includes some dehydrated vegetables that are used as seasonings
but are not included in the FDA definition of spices.
The 23 firms in the spice manufacturers group that; use EO
sterilization are distributed evenly across the country, except
for a large concentration of seven firms in California.10'13
Table 8-10 shows some summary statistics on the sterilization
chambers and sterilant gas used by these manufacturers.1 The
23 facilities operated a total of 27 sterilization chambers in
1986, an average of only 1.2 per facility. The number 6f
chambers per facility ranged from one to four. Average: chamber
volume per facility was 33.1 m3 (1,168.9 ft3), which isj
considerably lower than the average for medical device puppliers
of 40.1 m3 (1,416.1 ft3). In addition, the range of chamber
volumes per facility, 0.1 m3 to 177 m3
(3.5 ft3 to 6,25p ft3),
was narrower for the spice manufacturers. Both average; chamber
volume per facility and the range of chamber volumes pejr facility
for the spices group were roughly equal to the average range for
[
the Pharmaceuticals group. Table 8-10 presents the 1987 data for
SIC 2099. j
The spice manufacturers used 124 Mg (122 tons) of EO and
[
287 Mg (282 tons) of total gas during 1986. Annual EO Use per
facility averaged 5.4 Mg (5.3 tons), while total gas us£ per
facility averaged 12.5 Mg (12.3 tons). Average total gas use per
facility is roughly equal to average total gas use for jthe
Pharmaceuticals group, but average EO use per facility is about
half the average for pharmaceutical manufacturers. Thei inert-gas
8-22
-------
TABLE 8-9. 1987 PERFORMANCE DATA FOR SPICK MANUFACTURES
(SIC 2099)
Product data
Value of shipments (106 $)
1987
9,815.8
Industry data
Total employment (103 people)
Production workers (103 people)
Capital expenditures (106 $)
Value added (106 $)
58.1
40.9
248.0
5,201.1
8-23
-------
TABLE 8-10. SUMMARY STATISTICS ON STERILIZATION CHAMBERS
AND GASES USED BY 23 SPICE MANUFACTURERS, 1988
Sterilization chambers = 27
Number per facility
Chamber volume per facility, m^ ( ft*)
Ethyleoe oxide use » 124.0 Mg/yr
(122.0 tons)
Use per facility, Mg/yr (tons)
Total gas use - 286.6 Mg/yr (282.1 tons)
Use per facility, Mg/yr (tons)
Average
1.2
33.1 (1,168.9)
5.4 (5.3)
12.5 (12.0)
Standard
deviation
0.6
14.3 (499.4)
11.2(11.0)
14.0 (13.8)
Range
!' 1-4
0.1-177.0
(3.5 - 6,250.7)
<0.05 - 40.0
! (< 0.05-39.4)
[
<0.05 - 56.3
(< 0.05 -55.4)
8-24
-------
for tlie spice manufacturers is 1.31, which is comparable to
the 2.88 ratio for the medical device suppliers.
8.4.2.5 Museums and Libraries. According to the
1982 Census of Services, the United States has 1,909 non-
commercial museums and art galleries (SIC 8411).17 In addition,
the country had 31,524 public and private libraries in 1987. The
sterilization data base contains data for 11 museums and
2 libraries. Interestingly, 4 of the 13 are in Massachusetts and
none are in the Southeast or Northwest.
Museums and libraries fumigate books, documents, and other
artifacts with EO chiefly to control insect pests and mold.
Museum experts report that EO is "especially valuable for
treatment of books and archival documents, furs, textiles, and
furniture."19 However, EO has one significant drawback as an
artifact fumigant: it settles in rubber, leather, wood, and
other organic materials, making it necessary for EO-fumigated
artifacts to be aerated for up to a month before they are safe to
handle.19'20
It is recommended that all organic materials be fumigated
before they are introduced into a museum or library collection.19
However, telephone conversations with museum and library
conservators who use EO revealed a range of fumigation criteria.
Some conservators fumigate all new articles, while others
fumigate only those articles that fail a visual inspection or
that have suspect backgrounds, such as books that were kept in a
damp basement.21'22
Ethylene oxide has several substitutes as a fumigant in
museum and library use. One possible substitute is sulfuryl
fluoride, marketed under the trademark Vikane™. Sulfuryl
fluoride is not absorbed by organic materials and dissipates more
quickly than EO.20 Furthermore, the cost of sulfuryl fluoride
has generally been comparable to the cost of 12/88, based on
retail prices and the recommended doses of each sterilant.19
Sulfuryl fluoride, however, is toxic at high concentrations. It
is also corrosive to metals, making it an unacceptable
8-25
-------
alternative for artifacts that contain metal. In particular, it
could damage books with staple bindings.23 \
Conversion from EO to sulfuryl fluoride would entail some
startup costs. Some EO fumigation chambers would require
modifications to use sulfuryl fluoride. Because sulfuryl
fluoride is corrosive to metals, the vent pipes from the chamber
must be stainless steel; installing these new vent pipes would
represent a startup capital cost for museums/libraries converting
to sulfuryl fluoride use.23 I
Sulfuryl fluoride is registered with the U.S. Department of
Agriculture as a "restricted" pesticide and, therefore,! can only
be applied by a certified applicator. Certified applicators must
pass a test administered by the U.S. Department of Agriculture.
Ethylene oxide is not registered as a restricted pesticide and
can, therefore, be applied by anyone.24 The time and elf fort
involved in passing the certified applicator test would; represent
another startup cost for museums/libraries converting tb sulfuryl
fluoride. Finally, sulfuryl fluoride is registered for use in
fumigation chambers, and presently no regulations control
sulfuryl fluoride emissions.24 :
Other alternatives to EO fumigation include deep freezing,
CO2 fumigation, and vacuum treatment. Several European!
institutions have tested the freezing method and reported that
maintaining -18°C (-0.4°F) for 48 hours kills 100 percent of
insect life in all stages of the life cycle. Freezers ranging
from 0.9 to 1.1 m3 (31.8 to 40 ft3) in size are most commonly
used.19 The necessary freezing apparatus costs approximately
$3,000 to $4,000.22 Placing artifacts in a vacuum or fumigating
them with carbon dioxide also kills insect life.21'25 These
i
three methods, however, do not kill mold and fungi; therefore,
they are only partial substitutes for EO. !
Table 8-11 provides some summary statistics on the|
sterilization chambers and sterilant gas used by the Hi museums
and two libraries included in this study.1 Each of theke
13 facilities operated one sterilization chamber in 1988. These
chambers averaged 2.60 m3 (91.82 ft3) in volume and ranged from
• I
8-26 '
-------
TABLE 8-11. SUMMARY STATISTICS ON STERILIZATION CHAMBERS
AND GASES USED BY 13 MUSEUMS AND LIBRARIES, 1988
Sterilization chambers =13
Number per facility
Chamber volume per facility, nr* (ft*)
Ethylene oxide use = 0.20 Mg/yr
(0.20 tons/yr)
Use per facility, Mg/yr (tons))
Total gas use = 1.68 Mg/yr
(1.65 tons/yr)
Use per facility, Mg/yr (tons/yr)
Average
1.0.
2.6 (91.8)
•
<0.05
(<0.05)
0.1(0.1)
Standard
deviation
0.0
3.7(130.7)
<0.05
(<0.05)
0.1 (0.1)
Range
1 -1
0.5 - 13.7
(17.7 - 483.8)
<0.05 - 0.05
(<0.05 - 0.05)
< 0.05 -0.4
(< 0.05-0.4)
8-27
-------
0.5 m3 to 13.8 m3 (17.7 ft3 to 487.3 ft3). The average chamber
volume per facility for this group is roughly one-twentjieth the
average chamber volume per facility for the medical device
suppliers. Additionally, all of the gas use figures are much
lower than are the figures for medical device suppliers!.1
. *"
Durzng 1988, the museums and libraries group used a total of
0.20 Mg (0.20 tons) of EO and 1.67 Mg (1.64 tons) of total gas.
Medical devices suppliers used over 3,000 times as much EO and
i
over 1,500 times as much total gas as did the museums and
libraries group. In addition, gas use per cubic meter of chamber
volume was much lower for this group than for any other group.
Medical device suppliers used 14 times as much total gas and
22 times as much EO per cubic meter of chamber volume as did the
museums and libraries group. Furthermore, this latter group used
substantially less EO per cubic meter of chamber volume I than any
other industry group in this study—12.6 kg/m3/yr I
(0.78 Ib/ft3/yr), while all other industry groups each used over
200 kg/m3/yr (12.5 Ib/ft3/yr). The inert-gas ratio for] the
museums and libraries group was 7.40, which suggests that
facilities in this group used 12/88 exclusively.
8.4.2.6 Laboratories. Four of the firms being considered
for regulation are commercial laboratory rat and mice breeders,
which are classified as SIC Code 2790. Animal breeders1use EO to
sterilize heat-sensitive plastics, heat- and water-sensitive
electronic lab equipment, and prepackaged articles required for
breeding. These articles include books, phones, optical
equipment, meters, microscopes, and face masks.14
Some alternatives to EO sterilization for laboratory animal
breeding do exist, although each has disadvantages. Pressurized
steam damages plastics, electronic equipment, and other!heat-/
water-sensitive materials. In addition, steam does not!permeate
prepackaged materials as effectively as EO. Dry heat degrades
those heat-sensitive materials that cannot withstand temperatures
of 160° to 165°C (320° to 329°F). Another alternative,Iperacetic
acid, is not registered as a pesticide and is considered to be
carcinogenic. It is also corrosive to most metals and plastics
8-28 i
-------
and has poor penetrating properties. Glutaraldehyde requires at
least 10 hours immersion and cannot be used with prepackaged
materials. It poses a recontamination danger from hand
manipulations and air and cannot be used for materials that would
be damaged by water.14
Two noncommercial research organizations, SIC Code 8922,
have also been identified as operators of sterilization chambers.
These firms use EO in the same manner as the lab animal breeders
mentioned above but on a somewhat more limited scale. Ethylene
oxide alternatives and disadvantages mentioned for animal
breeders also apply here.14
The remaining facilities classified as laboratories are two
research and development labs (SIC Code 7391), one medical lab
(SIC Code 8071), and one commercial testing lab (SIC Code 7397).
These firms are considered together because their sterilization
practices are similar. In particular, these facilities
investigate whether medical devices can be effectively sterilized
and yet still function within the human body. Alternatives to EO
for these facilities are limited; EO must be used for heat- and
moisture-sensitive medical devices and drugs.14
Table 8-12 reports some summary statistics on the
sterilization chambers and sterilant gas used by the
10 laboratories included in this study.1 These facilities
operated 22 EO sterilization chambers during 1988, an average of
2.2 per facility. The number of chambers per facility ranged
from one to five. Average chamber volume per facility was
4.15 m3 (146.56 ft3) but varied from 0.05 m3 to 16.1 m3 (1.77 ft3
to 568.57 ft3) per facility. The laboratories group averaged
about the same number of chambers per facility as the medical
device suppliers group; however, the laboratories group had a
lower average chamber volume per facility than did all other
groups (except the museums and libraries group). This comparison
suggests that the laboratory group operated smaller chambers than
did the other groups (except the museums and libraries group).
The laboratory group used 9.55 Mg (9.40 tons) of EO and
78.1 Mg (76.9 tons) of total gas during 1988. The group average
8-29
-------
i
i
TABLE 8-12. SUMMARY STATISTICS ON STERILIZATION CHAMBERS
AND GASES USED BY 10 LABORATORIES, 1988
Sterilization chambers = 22
Number per facility
Chamber volume per facility, irr (fr)
Ethylene oxide use — 0.20 Mg/yr (0.20 tons)
Use per facility, Mg/yr (tons/yr)
Total gas use — 1.68 Mg/yr (1.658 tons/yr)
Use per facility, Mg/yr (tons/yr)
Average
2.2
4.15 (146.56)
0.96
(0.91)
7.8 (7.7)
Standard
deviation
1.6
4.81 (169.86)
2.57
(2.53)
21.5 (21.2)
Range
!
1 1-5
!. 0.5-16.1
! (17.66-568.57)
, <0.05 - 8.68
(<0.05-8.54)
'<
i <0.05 - 72.3
(< 0.05 -7 1.2)
8-30
-------
of O.96 Mg (O.94 tons) of EO per facility is roughly equal to the
spice manufacturer's use per facility but less than all other
groups, excluding museums and libraries. The laboratories also
used less total gas per facility, 7.8 Mg/yr (7.7 tons/yr), than
did all other groups excluding museums and libraries. The inert
gas ratio for laboratories was 7.17.
8.4.2.7 Contract Sterilizers. As mentioned earlier, a
subset of sterilization facilities sterilize products on a
contract basis. These contract facilities are normally
classified under SIC Code 7399 (business services, not elsewhere
classified). However, depending on the main type of product
sterilized, the facility may fall under another related category.
For example, one contract sterilizer works with surgical garments
and is classified under SIC Code 7218 (industrial launderers).
In addition to the facilities whose primary function is contract
sterilization, several facilities that sometimes accept contract
work are classified under a different category. These are
especially prevalent within the medical device industry.
Reports vary as to the number of contract sterilizers in the
United States. One official at the FDA estimated that there are
100 to 125 contract sterilizers of all types (i.e., EO, steam,
radiation, etc.). Of these, approximately 60 are EO
sterilizers.26 Another source at the FDA estimated that there
are 65 contract sterilizers of medical devices, 18 of which do
contract work only.27 Only 17 of the 188 facilities in this
study have been identified as contract sterilizers exclusively;
however, an undetermined number of the other facilities also
accept contract work.
The price for contract sterilization varies with the type of
sterilization performed. The prices for EO contract
sterilization are calculated based on the time the product
remains in the chambers (length of cycle) and the amount of gas
used. The pressure chambers contain a limited amount of space,
and the cost to the sterilization firm is the same whether the
chamber is completely full or not. Thus, the price per cubic
8-31
-------
foot of product sterilized also varies depending on howi full the
chamber is.
Some contract sterilizers said in telephone conversations
that the articles they processed tended to be of awkward shapes
and sizes, limiting the number of articles in the sterilization
chambers. Other firms processed smaller and denser loads. The
prices quoted were those for full chambers and averaged! about
$0.04 per-cubic meter ($1.30 per-cubic foot).28
Typically, sterilizing with 12/88 is more expensive than
sterilizing with pure EO. The difference in price between
sterilizing with pure EO and with 12/88 is due to the increased
amount of gas necessary to sterilize with 12/88.2 In spite of
the higher cost, facilities have used the 12/88 formulaltion
because some products or packaging cannot withstand sterilization
with pure EO, as was noted previously. In addition, pure EO is
explosive, making worker safety a concern. (Note: However,
during the early 1990's many facilities are switching to pure EO
because of chlorofluorocarbon (CFC) phase-out regulations.)
One expert on sterilization predicted 30 percent growth in
I1
the late 1980's and early 1990's for contract sterilizers.2
Growth would then continue at a rate of 4 to 5 percent annually.
Implicit in this projection is the assumption of no additional
Federal regulations in the future. Any new regulations would
probably augment the positive growth of contract sterilization,
because smaller facilities might cease in-house sterilization
operations and begin sending products offsite. [See '
Section 8.2.1 for further discussion of the possible substitution
of contract sterilization for in-house sterilization due to the
candidate NESHAP controls on EO emissions.] Indeed, this is
j ' i
already occurring because of CFC regulations. 1
Table 8-13 contains some summary statistics on thei
sterilization chambers and sterilant gas used by the 17 contract
sterilizers included in this study.1 These 17 facilities
operated 62 EO sterilization chambers in 1988—an average of 3.6
per facility, which is more than any other group in the study.
The contract sterilizers also had more chamber volume than any
8-32
-------
TABLE 8-13. SUMMARY STATISTICS ON STERILIZATION CHAMBERS
AND GASES USED BY 17 CONTRACT STERILIZERS, 1988
Sterilization chambers = 62
Number per facility
Chamber volume per facility, m3 (ft3)
Ethylene oxide use = 336.9 Mg/yr (331.6 tons)
Use per facility, Mg/yr (tons)
Total gas use = 1,567.6 Mg/yr
(1,542.9 tons/yr)
Use per facility, Mg/yr (tons/yr)
Average
3.6
87.1 (3,075.9)
19.8 (19.5)
92.2 (90.7)
Standard
deviation
2.8
72.6 (2,563.8)
13.0 (12.8)
95.8 (94.3)
Range
1-10
13.8 - 277.5
(487.3 - 9,799.8)
1.5 - 97.7
(1.5 - 96.2)
10.6 - 359.4
(10.4 - 353.7)
8-33
-------
other group, averaging 87.1 m3 (3,083.0 ft3) per facility.
During 1986, the contract sterilizers used 337 Mg (332 tons) of
EO and 1,568 Mg (1,543 tons) of total gas—somewhat more than
half the amounts used by the medical device suppliers. However,
the contract sterilizers used much more EO and total gas than any
other group on a per-facility basis. The inert-gas ratao for
contract sterilizers was 3.65, indicating that contract
sterilizers fall in the middle of the industry groups regarding
their reliance on pure EO. This is not surprising because
contract sterilizers service diverse industries with varying
sterilization requirements. |
8.5 DEMAND FOR ETHYLENE OXIDE STERILIZATION SERVICES
The demanders of EO sterilization services are the1
facilities that produce (or, in the case of museums and
1
libraries, acquire) the goods that require sterilizatio|n. The
demand for sterilization arises because sterilization is
necessary to ensure the ultimate safety and effectiveness of the
object sterilized. Specifically, many products cannot be
I
marketed unless they meet FDA sterilization standards. [ Medical
device suppliers sterilize their products because inadequately
sterilized products could cause harmful health effects Ifor users
of the devices. Similarly, spice manufacturers sterilize their
products because they may otherwise be damaged or contaminated by
insects, molds, or bacteria.
The demanders can be separated into two groups. One group
demands sterilization of the products they produce and also
satisfies their own demand by sterilizing the products in-house.
These are the facilities profiled in Sections 8.4.2.1 through
8.4.2.6. The other group of facilities demanding EO |
sterilization satisfies their demand by using the services of
contract sterilizers. We assume that these demanders are
facilities in the same industry groups as the facilities that
perform in-house EO sterilization. Rather than sterilize their
own goods in-house, however, they demand the services pf a
contract sterilizer. •
8-34
-------
The market for sterilization services may be viewed as being
in equilibrium in the absence of an air emission standard.
Demanders of sterilization compare the costs of sterilizing
products in-house with the costs of sending their products
offsite to a contract sterilizer. Likewise, they compare the
costs of the various types of sterilization that can be employed
with their products. They select the sterilization technique
that minimizes the cost of producing their good or service,
including the cost of sterilization. Imposing an air emission
standard on EO sterilizers will increase the cost of this type of
sterilization relative to other types. It will also increase the
cost of producing goods requiring sterilization. The following
section analyzes the impact of this relative increase in costs.
8.6 ECONOMIC EFFECTS OF CANDIDATE NESHAP CONTROLS UNDER THREE
CONTROL OPTIONS
This section analyzes the economic effects that the
candidate NESHAP controls will have on sterilization facilities
under three control options.29 As discussed above, EO has been
designated a probable human carcinogen, so EPA has developed
three possible control options representing increasing levels of
stringency. Imposing controls on EO sterilization will increase
the cost of performing this type of sterilization. This increase
in sterilization costs will, in turn, increase the cost of
producing goods and services in the industry groups that demand
sterilization.
In the following sections, the provisions of the three
control options are summarized. Then, the theoretical framework
for analyzing economic impacts that increase production costs are
described. Next, the analytical procedure used to evaluate the
impacts of the control options is described, and the empirical
results of the analysis are presented.
8.6.1 The Three Control Options
The three control options assessed in this analysis
represent increasing levels of stringency of control:
1. Option l controls only emissions from the chamber vent
and the vacuum pump drain;
8-35
-------
2. Option 2 controls these two sources of emissions, plus
emissions from the aeration room; and
3. Option 3 controls emissions from the main chamber vent,
I
the vacuum pump drain, the aeration room, and the rear chamber
i
exhaust. :
8.6.2 Theoretical Framework for Economic Impact Analysis
The ideal procedures for estimating the economic effects of
the proposed NESHAP controls use a framework based on the supply
and demand for goods or services in the regulated market. [For a
detailed description of this framework, see Chapter 2 in
Reference 29.] A market demand curve describes the maximum
quantity (per period) of a commodity, Q, that individuals or
firms are willing to purchase at various prices, ceterus paribus
(all else equal). As shown in Figure 8-1, demand curves slope
downward, indicating that consumers are willing to buy more of Q
at lower prices than at higher prices. This assumes that all
other factors that might influence demand—for example,;income,
prices of related goods, and tastes or preferences—do not
change.
If the market process establishes a price of Plf consumers
will purchase Q^ of the commodity for a total expenditure equal
to OP^BQ^. Because a demand curve measures maximum willingness
to pay for each unit of a commodity, the total willingness to pay
I
for Q! is the entire area DABO^—total expenditures plus the
triangle P-j^AB. This triangle, which is the difference between
what consumers actually pay and the amount they are willing to
pay, is known as consumer surplus. It is a good empirical
approximation of the dollar value of the well-being consumers
receive from consuming a commodity, over and above what|they pay
for the commodity.
The other principal construct in our conceptual framework is
a market supply curve. A supply curve shows the maximum output
(per time period) of a commodity that firms are willing
to supply
at various prices, ceterus paribus. The upward slope of supply
curves (as shown in Figure 8-2) indicates that firms are willing
to produce more at higher prices than at lower prices, assuming
8-36
-------
PRICE
($/Q)
CONSUMER SURPLUS
1 QUANTITY
(Q/TIME)
Figure 8-1. Demand curve for Commodity Q.
8-37
-------
PRICE
($/Q)
c
0
PRODUCER
SURPLUS
QUANTITY
(Q/TIME)
Figure 8-2. Supply curve for Commodity Q. j
8-38
-------
that other factors influencing the supply curve—prices of inputs
(such as labor, energy, and machinery) and production
technology—do not change.
If the market process establishes a price of P,, then
suppliers will produce Q^ units of the commodity, receiving
OP1BQ1 in total revenues. However, the cost of producing these
Q-L units is represented by the area OCEQ^. The triangle CP-j^B,
which is known as producer surplus, is the difference between the
minimum amount firms would accept for the Q1 units and the actual
amount they receive for these units. Producer surplus is a good
empirical approximation of the dollar value of the returns that
firms experience from producing a commodity, over and above the
costs of production.
Installing and operating controls on EO emissions from
sterilization chambers will increase the cost of sterilization.
In a demand-supply framework, this additional cost is represented
by an upward shift in the supply curve (from S± to S2 in
Figure 8-3). This upward shift in the supply curve leads to a
higher market price (P2) and a smaller quantity demanded (Q2).
The changes in price (from P^ to P2) and quantity (from Q1 to Q2)
are market adjustments attributable to the emissions controls.
The cost of this change in market-clearing price and
quantity due to the emissions controls is represented by the area
CDEB—the area between the two supply curves S^^ and S2 and under
the demand curve. This area, which constitutes the cost that
society experiences because of the emissions controls, equals the
sum of the additional cost of producing Q2 units of the commodity
(area CDEF) plus the foregone consumer and producer surplus
(Ql - Q2) on the units of the commodity that are no longer
produced or consumed (area EFB). Equivalently, the social cost
of emissions controls can be determined by aggregating the impact
of the controls on the well-being of consumers and producers of
affected commodities. In other words, social cost equals the sum
of the change in consumer and producer surplus as a result of the
price and quantity adjustments.
8-39
-------
PRICE
($/Q)
QUANTITY
(Q/TIME)
Figure 8-3. Market equilibrium with and without an upward shift
in the supply curve due to ethylene oxide emissions controls.
STL snows market supply witout EO emissions controls.
S2 shows market supply with emissions controls.j
8-40 !
-------
The simple model in Figure 8-3 ignores two potentially
important aspects of the economic impact of the candidate NESHAP
control under the three control options. First, the controls
will apply both to existing and new plants (when they are
constructed) that produce sterilized products. Thus, two supply
curves are relevant: a supply curve from existing plants and a
supply curve from new plants. The candidate NESHAP controls will
shift both of these supply curves upward, although the magnitude
of these shifts will probably differ. The potential effects of
the candidate NESHAP controls on new facilities are addressed in
Section 8.8.2.
Second, this discussion implies that all EO sterilization is
offered in a market. In fact, much EO sterilization is performed
in-house. In this case, the facility's demand for sterilization
services is met by its own supply of sterilization services.
These facilities do not generally offer to sterilize products for
others on a contract basis, so their services are not part of
market supply and have no market price. Only the EO
sterilization performed on a contract basis is actually marketed.
Therefore, employing the supply and demand framework to analyze
the impacts of the regulation is not possible. Rather, because
most sterilization takes place in the context of producing some
other good or service, the impact of the regulation on the
production of those goods and services has been analyzed.
As noted above, imposing air emission controls on EO
sterilization increases the cost of sterilization, which in turn
increases the cost of producing goods requiring sterilization.
The candidate NESHAP controls will probably cause some firms to
substitute contract sterilization for in-house sterilization. If
the candidate NESHAP controls are promulgated, facilities
currently sterilizing in-house using EO will face four possible
alternatives:
1. Adopt the candidate controls;
2. Switch to another sterilization process;
3. Discontinue in-house sterilization, switching to
contract sterilization; or
8-41
-------
4. Discontinue sterilizing or producing the sterilized
product. i
Undoubtedly, substituting contract sterilization for in-
house sterilization (Alternative 3) will be the least-cost
alternative for some facilities. Others may choose to switch to
another method of sterilization. Finally, some facilities may
decide to discontinue sterilization or to stop producing products
that require sterilization. Because of a lack of necessary data,
no attempts are made to decide which facilities will respond in
which way to the candidate NESHAP. [For a detailed explanation
of the substitution effect, see Section 8.7.4.] |
Contract sterilizers' costs will also be increased!by
i •*
imposing the control options. Their choices of possible
responses are limited to three alternatives: i
1. Adopt the candidate controls; |
2. Switch to another sterilization process; or i
3. Discontinue sterilization.
8.6.3 Analytical Procedure !
The theoretical framework for analyzing economic impacts
involves estimating changes in the market price and quantity sold
of a product or service. As noted above, however, in-house
sterilization is not marketed. Consequently, using the'supply
and demand framework to analyze the economic effects of!the
candidate commercial sterilization NESHAP is not possible. As an
i
alternative, EO sterilization is analyzed within the context of
the production of the goods and services requiring sterilization
i
and a more qualitative approach is used that approximates the
ideal approach. !
This approach has four parts. First, dividing commercial
sterilizers into seven industry groups imposes some homogeneity
on the facilities. Second, each facility's chamber volume and
annual EO use is used as proxies for the quantity of sterilized
goods produced. By dividing these proxies into the annual
compliance cost under the three control options, rough estimates
are obtained of the per-unit production cost increase caused by
the candidate NESHAP. Third, for each control option, dividing
8-42
-------
each facility's annual compliance cost by its total baseline cost
of sterilization produces an upper-bound for the percentage
increase in production cost caused by the NESHAP under each
control option. Finally, dividing each facility's annual
compliance cost under each control option by its total sales
produces a lower-bound estimate of the percentage production cost
increase. Each facility's actual percentage production cost
increase attributable to the candidate NESHAP under each control
option lies somewhere between its upper- and lower-bound
estimates.
8.6.3.1 Industry Grouping. The 188 affected facilities
were divided into seven industry groups based on their SIC Code:
medical device suppliers, other health-related manufacturers,
Pharmaceuticals manufacturers, spice manufacturers, museums and
libraries, laboratories, and contract sterilizers. The firms
within each of these subgroups produce a more homogeneous mix of
goods than does the aggregate group. Nevertheless, the product
mix is still quite diverse within the industry groups, as is
demonstrated in Section 8.1.
8.6.3.2 Chamber Volume as an Output Measure. The sum of
the volumes of all sterilization chambers at a facility is one
measure of the facility's sterilization capacity, if
sterilization cycles for all sizes and types of chambers are of
equal duration, then a facility with twice the chamber volume of
another facility also have twice the sterilization capacity. If
this assumption is not true, then the direct relationship will
not hold. For example, if larger chambers undergo longer cycles,
then chamber volume will overstate sterilization capacity for
large chambers and understate capacity for small chambers. If
small chambers undergo longer cycles, then the reverse will hold.
Chamber volume can be used not only as a measure of capacity
but also as a measure of output under the following two
additional assumptions.
1. All facilities within an industry group perform about
the same number of sterilization cycles per year; and
8-43
-------
2. The sterilization chambers are always filled With
products to approximately the same level. >
Under these assumptions, chamber volume would be highly
correlated with the volume of goods sterilized. !
How valid are these two assumptions? Data from the EPA
commercial sterilization data base indicates that facilities do
not perform equal numbers of sterilization cycles per chamber,
even within industry groups.1 These data indicate that; the first
assumption is not true. Therefore, chamber volume may be a poor
output measure. !
8.6.3.3 EO Use as a Measure of Output. Ethylene [oxide use
is the second surrogate output measure. The EPA commercial
sterilization data base provides data on the amount of [EO used
per facility during 1986 (1988 or 1985 for some facilities).1
Ethylene oxide use is a better output measure than chamber volume
because EO use varies directly with the volume of goods
sterilized, even if firms do not run a consistent number of
sterilization cycles per chamber. However, using EO as a measure
of the volume of products sterilized requires making the
following two assumptions: I
1. The concentration of EO per unit of chamber vjolume is
roughly the same for all chambers; and '
2. Sterilization chambers are filled with products to
roughly the same level for each cycle. j
Again, the assumptions necessary for EO use to be ja valid
measure of output are probably not met. First, the concentration
of EO per unit of chamber volume varies depending on the type of
gas mixture used. Sterilization with pure EO requires ja lower
concentration of EO than does sterilization using 12/88. Also,
although we have no supporting data, we can speculate that
sterilization chambers may not always be filled with products to
the same level. Probably facilities running frequent
sterilization cycles would always fill their chambers to capacity
to minimize costs. However, facilities that sterilize i
infrequently may run some cycles at less than capacity [because
they face sporadic orders for sterilized goods and shipment
8-44
-------
deadlines. To the extent that this occurs, EO use may overstate
the volume of goods sterilized by small-volume sterilizers. If
so, this overstatement of facilities' output would make the
industry supply curve appear flatter (i.e., more price-elastic)
than it actually is.
To this point, a potentially important problem with the
output measures has been ignored. As mentioned above, many types
of products are sterilized with EO. Chamber volume and EO use
are measures of the physical volume of products sterilized, but
with such a diverse group of products, volume is not an
appropriate measure of output. For example, one facility may
sterilize pacemakers while another sterilizes scalpels. Output
(in physical or monetary units) per cubic meter of chamber volume
will likely differ substantially between pacemakers and scalpels.
A similar conclusion seems reasonable regarding EO use as an
output measure. In summary, the diversity of products sterilized
in the seven industry groups undermines the usefulness of chamber
volume and EO use as proxies for facility output.
In addition to measures of total annualized compliance cost
(TAG) per unit of output, approximated by the measures described
above, two measures are computed that approximate TAG as a
percentage of the total baseline cost of producing sterilized
products. Ideally, TAG would be reported as a percentage of the
total cost of producing sterilized products. However, no data on
such costs at the affected facilities are available. Therefore,
TAG is reported as a percentage of the total annualized baseline
cost of sterilization at a facility (TAC/C) and TAG as a
percentage of total facility sales (TAC/S). Because the total
cost of producing sterilized products (hereafter referred to as
total production cost) equals or exceeds the sterilization cost
for any facility, TAC/C is an upper bound for TAC/total
production cost. Conversely, total production cost is generally
less than or equal to facility sales, particularly because most
affected facilities produce a mix of sterilized and unsterilized
products. Therefore, TAC/S is a lower bound for TAC/total
production cost. The denominators of these two measures,
8-45
-------
baseline sterilization costs and facility sales, are described in
the following two sections. !
8.6.3.4 Baseline Cost of Sterilization. The total annual
costs of sterilization for each facility were calculated using a
model chamber approach. Specifically, the capital cost and
operating cost per cycle are calculated for representative size
chambers in the following chamber categories: 12/88, pure EO,
nonflammable EO/CO2, and flammable EO/CO2> Four representative
chamber sizes for the 12/88 gas type were chosen, four for pure
EO, three for flammable EO/CO2, and two for nonflammable EO/CO2-
Representative chamber sizes were selected by first minimizing
i
the total variance between the representative chamber sizes and
the sizes of chambers currently in use, as reported in the EPA
commercial sterilization data base. Then the selection of
representative chamber sizes was finalized by reviewing! scatter
plots of actual chamber sizes and talking with a vendor of
sterilization chambers. i
A sterilization chamber vendor provided estimates of capital
and operating costs for each model chamber. Capital costs
include the cost of the sterilization chamber, the cost of
i'
installing the chamber, and the cost of explosion protection
equipment for use with flammable sterilant gases. Chamber
operating costs include labor, electricity, and sterilant gas
i
costs.
After determining the capital and operating costs for the
model chambers, the chambers were divided at the affected
facilities into the four gas types outlined above. The operating
cost per cycle and capital cost for each chamber were determined
by interpolating between the model chambers based on chamber
size. Multiplying the operating: cost per cycle by the annual
i
number of cycles for the chamber, as reported in the EPA
commercial sterilization data base, produced the annual' operating
cost of the chamber. The capital cost was annualized using a
I
useful life of 10 years and an interest rate of 10 percent, and
the annualized capital cost and the annual operating co;st were
added to obtain the annual sterilization cost for the chamber.
i,
8-46
-------
Next, determining the annual baseline cost of sterilization at
the facility involved aggregating the annual sterilization costs
for individual chambers at each facility. Then, for facilities
with controls in place at baseline, the annualized cost of these
controls was added. Finally, because the compliance costs are
given in last quarter 1987 dollars, the Producers' Price Indices
were used for all commodities for 1986 and 1987 to adjust the
baseline facility costs from 1986 dollars to 1987 dollars. The
costs, therefore, represent 1986 data in 1987 prices.17
8.6.3.5 Facility Sales. Total facility sales provide an
upper-bound estimate of the total costs of producing sterilized
goods for several reasons. First, total sales for a facility
include sales of sterilized and unsterilized products. For
facilities that sterilize only a small portion of output, total
sales may substantially exceed sales of sterilized products.
Second, sales reflect the price of final products multiplied by
the output of these products. They do not directly measure
production costs. In general, sales equal production costs plus
producer surplus. Producer surplus is only zero when the supply
curve is horizontal or the firm is the marginal firm in its
industry; therefore, using sales to measure production costs
implies one of these two conditions. If producer surplus is not
zero, then total sales will overstate total production costs.
As noted, sales are reported on a facility-specific basis.
When a facility did not report its own sales and the parent
company sales were available, the amount of sales per employee
for the parent company was multiplied by the number of employees
at the facility. This calculation assumes that the amount of
sales per employee at the parent company and the facility are the
same, which would not be true, for example, if the facility were
more capital-intensive than the parent company. No information
regarding the validity of this assumption is available.
Where neither sales nor employment information was
available, estimating facility sales was accomplished by using an
ordinary least-squares regression of facility sales on facility
EO use. Dummy variables were included to represent the seven
8-47
-------
industry groups, and a log-log functional form was usedl As with
baseline annual sterilization costs, facility sales were adjusted
from 1986 dollars to 1987 dollars, using the Producers'[Price
Indices for all commodities for 1986 and 1987.17 :
8.6.4 Results j
As explained in the previous section, the economic;analysis
was based on the annual cost of complying with the candidate
NESHAP controls under each of three control options. The three
control options represent increasing levels of stringency of
control, as described in Section 8.6.1. The control costs under
each control option are evaluated relative to chamber volume, EO
use, total baseline sterilization costs, and total facility
sales. The summary statistics on these values are presented in
this section. I
8.6.4.1 Total Annualized Cost (TAG). Total annualized cost
f
measures the annual engineering cost of compliance with\the
candidate NESHAP controls under each of the three control
I
options, as described in Chapter 7 of the Background Information
Document. Total annualized cost under Option 1 equals the
capital cost of an acid water scrubber and other necessary
hardware annualized over a 10-year period at an interest rate of
!
10 percent, plus annual operating costs for the scrubber,
including labor, materials, and ethylene glycol disposal costs.
Electric and water costs for scrubber operation were assumed to
be insignificant and are hot included. Emissions from the vacuum
pump drain were assumed to be controlled using a water-sealed
vacuum pump(s) with closed-loop recirculation and a liqijiid-gas
separator. Costs for these were estimated, as were costs for the
piping for manifolding the chambers at a facility to the existing
control device or to a scrubber.30 ;
The total annualized compliance costs under Option!2 include
the costs described under Option I plus the additional costs of
controlling the emissions from the chamber in which the product
is aerated. For the cost analysis, it was assumed that!a
facility could use insulated shipping containers as modular
aeration units, replacing all existing aeration processes
8-48 •!
-------
conducted in aeration rooms over 84 cubic meters. For these
facilities, the number of modular units needed was estimated as
were the costs for purchasing and installing these units. Costs
were assessed for both catalytic oxidation units and gas/solid
reactant systems that might be used to control the emissions, for
any manifolding required, and for materials, labor, and other
operating costs. The least costly approach, a gas/solid reactant
system, was selected for the impact analysis.31
Finally, the total annualized compliance cost under Option 3
includes the compliance costs under Option 2 plus, for those
facilities with rear chamber exhausts on some or all of their
chambers, the costs of controlling the emissions from that
source. Facilities with total sterilizer volumes less than 7 m3
probably do not have rear chamber exhausts and therefore were not
included in this cost analysis.32 Several control methods were
considered for rear chamber exhaust emissions, including
installing dedicated scrubbers and manifolding the rear chamber
exhaust emissions to the aeration room control. This impact
analysis used the least costly method, the dedicated add-on
scrubbers.
Table 8-14 reports the median and range of TAG per facility
for each of the seven industry groups under each control option
and includes only facilities incurring positive total annual
compliance costs. As shown in Table 8-15, under Option 1,
29 facilities do not incur compliance costs. Under Option 2,
only two facilities escape compliance costs, and under Option 3,
all facilities but one incur at least some compliance cost.
Under Option 1, the median TAC's range from $8,400 to
$44,000. The highest median TAG, $44,000, is incurred by
contract sterilizers. Spice manufacturers experience the second
highest median TAG, $35,000. Other industries with relatively
high TAC's are other health-related manufacturers (median
TAG of $31,900) and medical device suppliers (median TAG of
$28,000).
A medical device supplier incurs the maximum TAG of any
facility—$128,000. Other industry groups having individual
8-49
-------
(0
H
H
EH
CU
CO
a
CO
i
1
1
It!
.1 §
2 o
•5 >•.
8-50
-------
10
3*
TOTAL ANNU
DUSTRY GROl
Cd 55
> H
§1
§ U
§S
§«
^
< 0
w £|
H j
W tt
88
go
H^
ll
t*4
w £
w H
/V"*
< w
o
EH U
55 c^
o "•"
f £H
CO
in °
1-1 w
TABLE
COMPLIAIi
§
S
cs
1
^^
c
_o
!•
o
O
A
U
o
II
o
A
II
U
o
A
^^
£
o
U
0.
eb
3
C
o
o
o
p
o
^
0
«
I
1
Miscellaneous/]]
a
o
a
0
*"*
*
i
's
u
'o.
oo
-
QO
m
r>
C
£
3
1
1 Pharmaceuticals
S
o
a
0*1
«
1
"5
s
1
j
o
ve
o
VO
0
0\
2
jfl
a
1"
1
| Medical equipm
"
o
«
o
«••«
VO
e
"53
cS
2
o
M
O
^"
0
S
1
I
«s
8 fa
8-51
-------
facilities that experience high TAG'S include pharmaceutical
manufacturers, with a maximum TAG of $106,000, and contract
sterilizers, with a maximum TAG of $104,000. |
At the same time, facilities in six of the seven industry
groups do not incur compliance costs under Option l (see
Table 8-15). These are facilities that already have the required
controls in place. Under Option 1, the lowest positive^ TAG for
an affected facility is a medical device supplier, with a TAG of
$600. The lowest TAG for an affected contract sterilizer, on the
other hand, is a relatively high $26,400. <
Under Option 2, the median TAG ranges from $17,700| for
museums and libraries to $63,100 for contract sterilizers. Also
experiencing relatively high median TAG under Option 2 kre spice
manufacturers and other health-related manufacturers, each with a
median TAG of $47,100. Under this control option, the highest
TAG facility is again a medical device supplier, with a TAG of
$240,000. Contract sterilizers and other health-related
manufacturers also have facilities with TAG'S over $200^000. The
lowest cost-controlled facilities, each with a TAG of $9,300
under Option 2, are found in the Pharmaceuticals manufacturers
and medical device suppliers groups. !
Finally, under Option 3, facilities with a chamberjfitted
with rear chamber exhausts incur incremental control costs,
compared to Option 2. The highest median TAG, again for contract
sterilizers, rises to $89,500. Spice manufacturers and other
health-related manufacturers incur a median TAG under Option 3 of
$65,800 and $65,500, respectively. Museums and libraries, on the
other hand, incur a median TAG of only $17,700 under Option 3.
The maximum value for a TAG is incurred by a medical device
supplier: $271,700. Pharmaceuticals manufacturers, otlier
health-related manufacturers, and contract sterilizers also
contain facilities that incur a TAG over $200,000 under!Option 3.
The highest TAG experienced by a museum or library, on the other
hand, is only $57,500 even under Option 3. j
8.6.4.2 TAG Relative to Chamber Volume. Table 8-16 reports
the median and range of total annualized costs per cubid meter of
. i
8-52 i
-------
Cn
O
u
H
i
(TAG
DUSTR
CO
CE
H
E
PLIA
AFFEC
ss
pq S
^s
*• .
><
E-i
H
S
vo
I
CO
CO
'I
s
1
I-?
5
£
7
r
a •
w »
*«
25
c.
e
-a
8?
3
°! e>
O '
6 =>
e
m O
s-
•* ee
W «
* "T.
o
e5
"I e>
O '
0 =>
§
u
^ vi
<= o
-a
s
I
•c
s
1
•e
1
1
-------
chamber volume per facility (TAC/CV) under each of the control
options. Under all three control options, the highest medians
are experienced by the museums and libraries group of facilities,
while the laboratories industry group contains the highest single
value of TAC/CV. These two industries incur relativelyilow
compliance costs, but they also have extremely low chamber
volume. Thus, their TAC/CV is high. Under Option 1, museums and
libraries have a median TAC/CV of $8,200/m3 ($289,581/ft3), while
laboratory facilities incur a median TAC/CV of $6,400/m3:
($2/260,142/ft3). Four of the other five industry groups
experience median values for TAC/CV of $2,200/m3 ($77,692/ft3) or
less. The maximum TAC/CV under Option 1 is experienced by a
medical device supplier: $297,700/m3 ($10,513,194/ft3)J
Under Options 2 and 3, the median TAC/CV for the museums and
libraries group increases to $17,400/m3 ($6l4,476/ft3), iwhile the
median TAC/CV experienced by the laboratories group is $10,400/m3
($367,273/ft3). The highest single TAC/CV is again experienced
by a medical device supplier: $626,100/m3 ($22,107,018/ft3).
Finally, under Option 3, the museums and libraries group again
experiences a median of $17,400 ($614,476/ft3) and the maximum
again is $626,100 ($22,110,549/ft3), experienced by a medical
device supplier. I
8.6.4.3 TAG Relative to EO Use. Table 8-17 shows |summary
statistics for total annualized compliance cost per metric ton of
facility EO use (TAC/EO) under each of the three control options.
As with TAC/CV, the museums and libraries and the laboratories
incur the largest impacts, when measured by TAC/EO. As 'with
TAC/CV, these facilities' relatively low compliance costs combine
with extremely low EO use to yield high TAC/EO values. Under
Option 1, the median TAC/EO incurred by the museum and iibrary
facilities is $l,216,100/Mg ($l,196,894/ton). The median
experienced by the laboratories group of facilities is
$163,100/Mg ($160,524/ton) . The other five industry gro'ups
experience much lower median TAC/EO, ranging from $8,100|/Mg
($7,972/ton) for contract sterilizers to $31,100/Mg ($30,609/ton)
for spice manufacturers. The maximum TAC/EO experienced by any
8-54 !
-------
ol
Hn
OS 10
EH Oi
w d>
£ O
pj o
U
& X
EH Q
CO Q
O W
°S
H U
U fa
8 .
H
|4
1
a-
•
1
&
ftt
"*. <"J
d o
.8
r<
-s-
~-
s
u
Ov
Q,
00 vo"
c< vn
V) ~
S
•J
i
u.
«
i;
I
s
2
i
r^
oo
i
M.
.s
I
IS
1^
ON *>
VQ ^
S
fj!
u u
li
S. c
i g
? s.
S o
H o
II
8-55
-------
facility under Option 1 is $9,152,000/Mg ($9,o07,458/ton),
experienced by a museum and library facility. Other industry
groups reveal maximum TAC/EO values ranging from $20,900/Mg
($20,576/ton) for contract sterilizers to $3,660,800/Mg !
(3,604,048/ton) for other health-related manufacturing. ' The
minimum TAC/EO is $0 for six of the seven industry groups under
Option 1 but is $183,000/Mg ($180,110/tonj for the museums and
libraries group.
Under Options 2 and 3, the same basic pattern is evident.
The highest median TAC/EO is experienced by the museums 'and
libraries industry, $2/352,800/Mg ($2,315,641/ton) under both
control options. The second highest median TAC/EO under Options
2 and 3 is experienced by the laboratories industry group:
$240,000/Mg ($236,210/ton) under Option 2 and $292,700/Mg
($288,077/ton) under Option 3. The maximum TAC/EO value for any
facility is experienced by a museum and library facility,
$19,382,000/Mg ($19,075,890/ton) under both control options.
Maximum TAC/EO values for other industry groups range frjom
$39,100/Mg ($35,530/ton) for contract sterilizers to !
$7,752,800/Mg ($7,630.356/ton) for other health-related
manufacturing groups under Option 2 and from $54,400/Mg !
($53,541/ton) for contract sterilizers to $7,752,800/Mgj
($7,630,356/ton) for other health-related manufacturing groups
under Option 3. |
8.6.4.4 TAC Relative to Baseline Annual Sterilization Cost.
i
Table 8-18 reports the medians and ranges of total annuailized
compliance cost as a percentage of baseline annual sterilization
costs (TAC/C) under the three control options. As mentioned
above, TAC/C represents an upper-bound estimate of TAC Relative
to total production cost, the desired impact measure. '
Under Option l, the median values for TAC/C range from
i
16.0 percent for medical device suppliers to 48.3 percent for
museums and libraries. The second highest median TAC/C jfor
Option 1 is 44.7 percent, experienced by spice manufacturers.
The laboratories industry group has a median TAC/C of !
27.6 percent, while contract sterilizers have a median 1JAC/C of
8-56 !
-------
o
W
w
~s
W
I
a
s
2
e
1
8
S3
P
00
00
5-
"•5
I
related
Miscel
liert
Medical
nd librari
Muse
•33-i ay
.SH
- =5 S 5
g.
« a-
8-57
-------
20.2 percent under Option 1. other health-related manufacturers
have a median TAC/C of 20.4 percent, while contract sterilizers
have a median TAC/C of 16.3 percent. Although some facilities in
six of the seven industry groups experience no compliance cost
under Option 1, other facilities experience TAC/C as high as
118.1 percent. This maximum TAC/C is incurred by a spice
manufacturing facility, other facilities with TAC/C over
100 percent are found in the Pharmaceuticals manufacturing
industry (maximum TAC/C of 117.2 percent) and museums and
libraries (maximum TAC/C of 115.8 percent).
Under Options 2 and 3, the median facility in the museums
and libraries industry experiences sterilization costs koubled by
the controls (TAC/C of 100.4 percent). The median TAC/C for the
other six industry groups ranges from 10.9 percent for Contract
sterilizers to 52.3 percent for spice manufacturers under Option
2, and from 13.8 percent for contract sterilizers to 68.4 percent
for spice manufacturers under Option 3. \
The highest TAC/C for any facility under Option 2 jis
166.7 percent, experienced by a Pharmaceuticals manufacjturer.
Other high TAC/C facilities are found in the museums and
libraries group (152.4 percent) and the spice manufacturing group
I
(144.9 percent). Even in the contract sterilizer industry, one
facility experiences a TAC/C of 63.4 percent. Under Option 3,
the maximum impacts range from 91.8 percent for contract
sterilizers to 225.8 percent for museums and libraries,i and four
of the seven industry groups have facilities that experience
177 percent TAC/C values, or higher. The lowest TAC/C facilities
in six of the seven industry groups incur TAC/C less than
8 percent under Options 2 and 3, but the lowest TAC/C facility in
the museums and libraries group has a value of 24.1 percent under
both Options 2 and 3. j
The relatively low TAC/CV, TAC/EO, and TAC/C for the
contract sterilizers under all three control options support the
hypothesis that the contract sterilizers may experience! economies
of scale in controlling EO emissions. In other words, the
candidate NESHAP controls may cause lower per-unit increases in
8-58 '
-------
production costs for contract sterilizers than for facilities in
the other groups.
8.6.4.5 TAG Relative to Total Facility Sales. Table 8-19
reports the total annualized compliance cost as a percentage of
total annual facility sales (TAC/S) under the three control
options. As described above, this measure represents the lower-
bound estimate of total annualized compliance cost as a
percentage of total annual baseline production cost.
Under Option 1, the median TAC/S values range from less than
0.1 percent (for museums and libraries and Pharmaceuticals
manufacturers) to 2.0 percent for contract sterilizers. The
highest TAC/S under Option 1 is experienced by a contract
sterilizer: 12.8 percent. Medical equipment suppliers have the
next highest maximum TAC/S under Option 1. One medical device
supplier experiences a TAC/S of 5.5 percent. The maximum TAC/S
for the other industry groups ranges from 1.6 percent, for the
other-health-related-and-miscellaneous-and-libraries groups, to
4.2 percent for museums and libraries.
Under Option 2, the contract sterilizers have a median TAC/S
of 3.9 percent, while all other industry groups experience median
TAC/S values between 0.0 percent and 0,5 percent. The maximum
TAC/S incurred under Option 2 is again a contract sterilizer.
For this facility, TAG represents 25.9 percent of annual sales
under Option 2. The maximum TAC/S values experienced by other
industry groups range from 2.3 percent, for other health-related
manufacturers, to 8.8 percent, for museums and libraries.
Under Option 3, the median TAC/S for contract sterilizers
rises to 4.6 percent. Again, all the other industry groups have
median TAC/S values less than one percent. The highest maximum
value is again a contract sterilizer, which incurs TAC/S of
29.7 percent. Industry maximums for TAC/S for the other six
industry groups range from 3.0 percent to 9.6 percent.
8.7 EFFECTS OF THE REGULATION ON SMALL BUSINESSES
8.7.1 Requirements of the Regulatory Flexibility Act
The Regulatory Flexibility Act requires the awareness and
consideration of small entities as regulations are being
8-59
-------
0
w
CS
r^
IH
O
w^
f^o
g
is
o
(s|
g
s.
"S
.s
£•
SS^SM
•
f
i
O
o
^
0
(S|
1
o
o
o
\o
t
o
o
o
M
u
1 Miscellaneoui/libontorii
o
CO
1
0
o
t-
o
*o
1
o
o
"1
o
r-
ts
i
o
o
o
e
o
.§
o.
to
>>•
o
0
o
«0
0
o
d
S
i
o
o
o
0
E
1 Pharmaceuticals manufac
e
<•>
o
o
o
(SJ
1
o
0
d
3
1
o
d
d
1 Other health related
>o
o>
i
o
d
o
0;
>|o :
o S.S § g ;
*rf WU o -» jo
Jfefi 0 n i5 i
="•=.«„ '
o S S | J
| 1 g -S -a
£ * ° u a
^ -o *> o 2
•S, i « g - ' !
i ! •§ 8 s
S S 5 S -S
aSffS'l !
Illi? !
" •? 8* a " i
fflll !
"5| | | 2
fllil !
8-60
i
-------
developed. The RFA requires a determination of whether there is
a "significant economic impact" on a "substantial number" of
small entities. The EPA has issued RFA guidelines containing the
following criteria for use in determining what is a significant
economic impact:
1. Annualized compliance costs increase total cost of
production by more than 5 percent;
2. Compliance costs as a percentage of sales for small
plants are at least 10 percentage points higher than for large
plants;
3. Capital costs of compliance represent a significant
portion of capital available to small entities; and
4. The requirements of the regulation are likely to result
in closures of small entities.
Normally, a substantial number of small entities are said to
incur significant impacts, if at least 20 percent of the small
entities experiencing increased costs as a result of the
regulation meet the above criteria. However, even if 20 percent
of affected small entities meet the above criteria, if that 20
percent represents only a very small absolute number of affected
entities, a substantial number of affected small entities do not
incur significant impacts.
8.7.2 Small Businesses Performing Ethylene Oxide Sterilization
Because EO sterilization is the major line of business for
contract sterilizers, these firms will probably incur relatively
large impacts. In the other industry groups, EO sterilization is
only one of many operations performed in the course of producing
another good or service. For most of these firms, EO
sterilization represents a small share of their total production
costs. Therefore, firms in these sectors are not expected to
incur significant economic impacts. In addition, in many of the
other sectors, firms performing EO sterilization are larger than
those in the contract sterilizer sector. For these reasons, the
contract sterilizer sector's impacts and firm sizes were first
examined in detail. Then, small business impacts were considered
in the other industry groups.
8-61
-------
8.7.3 Small Businesses in the Contract Sterilizer Industry Group
i
Firms in the contract sterilizer sector are considered small
i
if they have annual sales of less than $3.5 million. Nine of the
17 facilities included in the contract sterilizer industry group
were contacted to obtain information on facility sales and credit
availability and cost, to characterize the contract sterilizers'
clients, and to investigate the possibility of substituting other
sterilization techniques for EO sterilization. Of the nine,
eight responded, with one of the eight indicating that it no
longer used EO. Additional information about firm sales was
obtained from Dun and Bradstreet's "Dun's Market Identifiers."
As expected, many of the contract sterilizers are small.
Based on sales information obtained from the facilities I or from
Dun and Bradstreet where available, and on our sales estimates
when no data are available from the other sources, 12 of the
remaining 16 facilities in the contract sterilizer sector still
performing EO sterilization are small. To estimate the[increase
in total production costs for these small businesses, baseline
i
sterilization costs were used as a proxy for total production
cost. This yields a conservative estimate of the totalj
production cost increase. Under control Option 1, eight of the
12 small businesses are expected to incur compliance costs
i
exceeding 5 percent of baseline sterilization costs. Under
control Option 2 and 3, 11 of 12 are expected to incur compliance
costs exceeding five percent of baseline sterilization costs.
Initially, therefore, a substantial number of small entities
may be significantly affected by the regulation. In talking to
the facilities, however, several things were discoveredjthat will
mitigate the severity of the impacts. First, some of the
facilities contacted indicated that EO sterilization isionly a
part of their business. Several mentioned that they also offer
other types of sterilization and that they are encouraging their
customers to substitute these other types for EO wherever
possible. Thus, even if the facilities stopped offering EO
sterilization, they might not close. Secondly, and potentially
more importantly, six of the eight facilities contacted;indicated
8-62
-------
that if their client industries were also regulated, they
expected demand for their services to increase. Their estimates
for the increases in their business ranged from 20 percent to
200 percent.
Because of the relatively lower per-unit compliance costs
incurred by contract sterilizers, some facilities currently
sterilizing in-house will probably choose, as a result of the
regulation, to stop sterilizing in-house and substitute contract
sterilization. If such a substitution occurs, then both the
revenues and the costs of contract sterilizers may increase as a
result of the regulation, and revenues may increase by more than
costs. The following section discusses substituting contract for
in-house sterilization in more detail.
8.7.4 Substitution of Contract Sterilization for In-House
Sterilization
As discussed above, EO sterilization is performed in-house
by facilities that specialize in producing other goods or
services, of which sterilization is a small but necessary part,
and by contract sterilizers who specialize in sterilizing goods
for other producers. In this analysis, in-house sterilization is
performed by medical device suppliers, spice manufacturers,
Pharmaceuticals manufacturers, other health-related
manufacturers, laboratories, and museums and libraries.
The cost of using contract sterilization includes some
additional costs not experienced in in-house sterilization.
These additional costs include the cost of transporting the
products to and from the contract sterilizer, the inventory cost
of products while in transit, the reliability and negotiation
costs of dealing with an outside supplier, and the cost of
products damaged or not properly sterilized. These are referred
to as transactions costs. The per-unit cost of actually
performing the sterilization is expected to be lower for contract
sterilizers because their higher volume enables them to take
advantage of economies of scale.
Figure 8-4 shows the market for contract sterilization prior
to the regulation. Because in-house sterilization is demanded by
8-63
-------
$/UNIT
STERILIZED
SCMI
DCMI
Qci QUANTITY OF GOODS
STERILIZED PER YEAR
Figure 8-4.
The market for contract sterilization without the
air emission standard in place. j
8-64
-------
the same facility that performs it, in-house sterilization
services are not marketed, so in-house services have no market
price. Facilities that produce goods that need sterilization
compare the cost of sterilizing them in-house with the cost of
using a contract sterilizer, including the transactions costs
described above. Without the air emissions standard in place,
the market for contract sterilization services is in equilibrium
at a price P±. At this price, producers of QC1 goods requiring
sterilization have found that the cost of sterilizing them in-
house exceeds the cost of using a contract sterilizer, including
the transactions costs, and have chosen to substitute contract
sterilization for in-house sterilization.
Imposing emissions controls on EO sterilizers will increase
the marginal costs of both in-house and contract sterilization.
Figure 8-5 shows the effect of the regulation on the
sterilization supply curves of a typical contract sterilizer and
a typical in-house sterilizer. These supply curves show the
relationship between the marginal cost of sterilization and the
number of goods sterilized per unit of time. The upward slope
indicates that, as more goods are sterilized per unit of time,
the cost of sterilizing each additional good (the marginal cost)
increases.
Marginal cost curve MCC1 shows the contract sterilizer's
marginal cost of sterilizing each quantity of goods without the
air emission standard in place, while MCC2 shows the marginal
cost of sterilizing each quantity of goods with the air emission
standard in place. The vertical distance between the two curves
is the total annual compliance cost per unit sterilized
(TAC/unit) for each quantity sterilized. For example, suppose
the contract sterilizer sterilizes Qc quantity of goods. (This
quantity is arbitrarily chosen; without information about the
market price of contract sterilization, it is not known what
quantity of goods the facility sterilizes.) If this sterilizer
processes Qc quantity of goods, the marginal cost of
sterilization without the regulation in effect is OA. The
marginal cost of sterilization at output level Qc with the
8-65
-------
$/UNIT
STERILIZED
$/UNIT
STERILIZED
B
A
Qc QUANTITY OF
GOODS STERILIZED
PER YEAR
QUANTITY OF
GOODS STERILIZED
PER i YEAR
A CONTACT STERILIZER
AN IN-HOUSE STERILIZER
Figure 8-5. Marginal cost curves for a contract sterilizer and
an in-house sterilizer, with and without the air emission
standard in effect. i
8-66
-------
regulation in effect is OB. The distance BA measures the
TAC/unit for .that quantity sterilized. Similarly, if the in-
house sterilizer sterilizes Qj quantity of goods (again, an
arbitrarily selected quantity), OC measures the marginal cost of
sterilization without the regulation in effect, OD measures the
marginal cost of sterilization at output level Qj with the
regulation in place, and DC measures the TAC/unit for that
quantity sterilized.
The compliance cost per Mg of EO used should be lower for
the contract sterilizers, on average, than for in-house
sterilizers. Because of economies of scale in controlling the
emissions, larger emitters, such as the contract sterilizers,
will incur lower TAG per unit of product sterilized. This
expectation is embodied in the relatively larger upward shift in
the supply curve for the in-house sterilizer (DC is greater than
AB). The economies of scale in compliance are reflected in our
estimated costs of compliance. As shown in Tables 8-17 and 8-18,
contract sterilizers have the lowest median TAG per Mg of EO used
and the lowest median TAG per dollar of annual sterilization
cost.
If nothing else changes, the air emission standard results
in a higher equilibrium price and a smaller total quantity of
contract sterilization performed. However, the demand for
contract sterilization is expected to change. Because of their
higher per-unit TAG, some of the in-house sterilizers will find
that their cost of sterilizing in-house, with the compliance
costs imposed, now exceeds the cost of using a contract
sterilizer. This condition will increase the market demand for
contract sterilization. At baseline, 16 contract sterilizers use
approximately 335 Mg (330 tons) of EO per year, and approximately
170 in-house sterilizers use about 1,482 Mg (1,459 tons) of EO
per year. If the quantity of EO used is closely related to
quantity of goods sterilized, then in-house sterilization
accounts for approximately 81 percent of the total quantity of
goods sterilized annually. Depending on the cost functions and
compliance costs of the individual in-house sterilizers, many of
8-67
-------
them may choose to substitute contract for in-house
sterilization, resulting in a substantial increase in the market
demand for contract sterilization.
Figure 8-6 shows a new equilibrium in the contract
sterilization market, with the market supply curve, SCH2> shifted
upward by the per-unit TAG of the regulation. The market demand
curve, DCJJJ, is shifted outward as a result of former in-house
sterilizers who have decided to substitute contract sterilization
!.
for in-house sterilization. In this figure, both the price and
quantity of contract sterilization have increased.
This absolute increase in contract sterilization may or may
not occur, depending on the actual positions and shapes ''of the
supply and demand curves. Contract sterilization's share of
total sterilization will definitely increase, however, because
their lower average TAG per unit sterilized will cause some in-
house sterilizers to decide to switch to contract sterilization.
In this example, contract sterilizers' revenues increase! bY more
than their compliance costs as a result of the regulation.
Revenues without the regulation in place are shown by the
rectangle OP1AQC1, and revenues with the regulation in effect are
shown by OP2CQC2. The change in revenues is shown by the
difference in these two rectangles, shaded in on the graph. The
TAC/unit is the vertical distance between the two market1 supply
curves, CB, so the compliance costs as a result of the rjegulation
are shown in the rectangle, DP2CB. The change in revenues in
this case exceeds the compliance costs for the contract
sterilization sector. :
In summary, adopting the emissions controls on EO :
sterilization will increase the unit cost of sterilization for
both in-house sterilizers and contract sterilizers. Because of
economies of scale in controlling emissions, the TAG per! unit
sterilized will be higher for the in-house sterilizers than for
the contract sterilizers. In the new market equilibrium!, some
in-house sterilizers will probably decide to substitute contract
sterilization for their in-house sterilization. The share of
contract sterilization in the sterilization market will increase
8-68
-------
$/UNIT
STERILIZED
SCM2
SCMI
DCM2
QUANTITY OF
GOODS STERILIZED
PER YEAR
Figure 8-6.
The market for contract sterilization with the air
emission standard in effect.
8-69
-------
as a result of the regulation, and the.absolute quantity
sterilized may actually increase. In fact, the market price may
increase by more than the TAG per unit sterilized for cbntract
sterilizers so that their revenues will increase by more than
their costs as a result of the regulation. For these reasons, a
regulatory flexibility analysis for contract sterilizers, a more
detailed examination of small entity impacts, is not required.
8.7.5 Small Business Impacts in Other Industry Groups '•
Many of the other industry groups in our analysis include
SIC codes for which the actual size definitions in the Federal
Register, 13 CFR Part 121, specify numbers of employees, rather
than sales. Unfortunately, no employee data for any of|the firms
are in the data base. Therefore, the sales criterion usjjed for
the contract sterilizers was applied. The results are shown in
Table 8-20. i
i
Of the approximately 170 facilities in the other industry
groups, only 22 are defined as small, based on their estimated
sales revenues. Using facility sales as a proxy for totlal
production costs, the number of these facilities that would
experience an increase in total production costs of more than
5 percent was estimated. Given this criteria for a significant
impact, l of the 22 experiences significant impacts under Option
1, 6 of the 22 under Option 2, and 8 of the 22 under Option 3.
Another factor to consider when examining the impacts of these
small facilities is whether a facility is a major or an larea
source. If a facility is an area source, it is possible that the
facility will be required to adopt less stringent controls than
major sources or may even be exempt from the regulation.; For
this regulation, an area source is a facility that emits less
than 10 tons of EO per year. Given this definition, 20 jof the 22
small businesses are classified as an area source. As a result
of the information presented above and because only a small
absolute number of small entities in each industry experience
significant impacts as a result of the regulation, it is1
concluded that a regulatory flexibility analysis is not Required.
8-70
-------
TABLE 8-20. SMALL BUSINESSES IN THE INDUSTRY GROUPS
PERFORMING IN-HOUSE STERILIZATION
Industry group
Laboratories
Spice manufacturers
Pharmaceuticals manu-
facturers
Other health-related
Med. device suppliers
Museums and libraries
No. of
small
entities
2
3
2
5
6
4
No. of significant impacts
Option 1
0
0
0
0
1
0
Option 2
1
1
1
0
2
1
Option 3
1
1
1
0
3
2
8-71
-------
8.7.6 Summary !
The majority of contract sterilizers are small businesses,
and most of the small businesses in the contract sector will
experience compliance costs exceeding 5 percent of baseline
sterilization costs. Because contract sterilizers, on average,
sterilize larger quantities than in-house sterilizers, they will,
on average, experience economies of scale in controlling their
emissions. This condition will result in lower total annual
compliance costs per unit sterilized for contract sterilizers
than for in-house sterilizers, which in turn will result in
contract sterilizers having a larger share of the sterilization
market with the regulation in effect than they have at baseline.
Depending on the extent to which supply and demand for Contract
sterilization change in response to the regulation, andjthe
characteristics of the supply and demand relationships,;contract
sterilizers may find that their revenues increase by more than
their costs as a result of the regulation. Finally, depending on
the positions and shapes of the supply and demand curves, the
absolute quantity of contract sterilization performed may
actually increase as a result of the regulation. In the other
industry groups, only a small absolute number of small firms is
significantly affected by the regulation. For these reasons, it
is concluded that a regulatory flexibility analysis is not
required. |
8.8 CONCLUSIONS ;
In this section conclusions are summarized regarding the
economic effects of the candidate NESHAP controls, under each
control option, on existing and new facilities.
8.8.1 Effects on Existing Facilities
This study suggests that the candidate NESHAP controls will
increase the cost of performing EO sterilization at affected
facilities. The incremental costs involved in moving from one
control option to the next appear to be substantial, on average.
As described at the beginning of this section, the increased cost
of sterilization (resulting in shifting the supply curve for
sterilized products upward) will probably result in reductions in
8-72
-------
the quantities produced of products sterilized using EO and an
increase in their prices. Because the data required to estimate
the elasticities of supply and demand for the products sterilized
using EO are not available, it is not possible to make
quantitative estimates of these quantity and price changes.
Faced with the increased costs of sterilization with EO,
facilities may choose one of the following four alternatives:
1. Pay the increased costs, implement the controls, and
continue to sterilize using EO;
2. Switch from EO sterilization to another onsite
sterilization process;
3. Discontinue onsite sterilization and switch to contract
sterilization; or
4. Discontinue producing the sterilized product.
Without quantitative estimates of the changes in price and
quantity for sterilized products, and without adequate data about
the relative costs of the alternatives facing the facility, it is
not possible to determine which alternative facilities will
choose. However, some facilities will probably decide to
discontinue EO sterilization as a result of the candidate NESHAP
controls.
Based on the median values reported in Section 8.6.4, the
candidate NESHAP is likely to significantly affect three industry
groups: museums and libraries, spice manufacturers, and contract
sterilizers. The candidate NESHAP will have a substantial effect
on all facilities in the museums and libraries group. These
facilities have very small chambers and use a very small quantity
of EO. Consequently, the TAC/CV and TAC/EO values for museums
and libraries are relatively high under all three control
options. The TAC/C values indicate that the candidate NESHAP
controls will increase sterilization costs by more than
50 percent for many facilities under Option 1 and will more than
double sterilization costs for many museums and libraries under
Options 2 and 3. Furthermore, the TAC/S values indicate that the
costs of the candidate NESHAP controls represent a
significant percentage of the operating budget for many of the
8-73
-------
museums and libraries. To the extent that museums and libraries
do not have ready access to capital markets, they may also have
difficulty getting the capital needed to acquire the relevant
control devices.
Museums and libraries do, however, have possible substitutes
for EO. As discussed in Section 8.1, sulfuryl fluoride] marketed
under the trade name Vikane™, may be an acceptable substitute for
EO in sterilizing nonmetallic objects. Converting to Vikane as a
sterilant involves some costs, including those of chamber
modification and training the operators to be certified ito apply
Vikane1*, a registered pesticide, other possible substitutes
include deep freezing, CO2 fumigation, and vacuum treatment.
Converting from EO to any of these would entail some conversion
costs. Conversion costs must be compared with the costs of
implementing the candidate NESHAP controls and the costs of
i
switching from onsite sterilization to using a contract i
sterilizer. Depending on the relative costs, some facilities may
choose to continue onsite sterilization, others may switch to
Vikane, and still others may use a contract sterilizer, i
Alternatively, some museums and libraries may decide to |use the
services of a contract sterilizer or discontinue fumigation
altogether, because of a lack of the capital needed either to
implement the candidate NESHAP controls or to switch to 1 Vikane™.
Also, some facilities may conclude that the value they receive
from fumigation does not justify the additional cost of ;
implementing controls, employing a contract sterilizer, jor
converting to Vikane™. j
The TAC/CV and TAC/EO are also fairly large for thel
facilities in the laboratories group, because of the beiow-
average size of their sterilization chambers and relatively small
quantity of EO they use. Nevertheless, the TAC/S and TA€/C
values for this group are low, suggesting that sterilization
i
costs are a very small part of total production costs at these
facilities. Without knowing the precise figure, an animal-
breeding laboratory indicated that sterilization costs are
"surely less than 1 percent of total production cost."33
8-74 !
-------
The,TAG/C median values for the medical device suppliers,
other health-related suppliers, and pharmaceutical manufacturers
range from 16.0 percent to 20.4 percent under Option 1, from
16.6 percent to 31.6 percent under Option 2, and from
24.4 percent to 40.1 percent under Option 3. Thus, the candidate
NESHAP controls will substantially increase sterilization costs
in the industry groups. However, sterilization costs are
generally very small relative to the total cost of producing
sterilized products in these industries. For example, a
pharmaceutical manufacturer estimated that sterilization costs
represent only about 3 percent of total production costs.8
Consequently, the candidate NESHAP controls probably will not
significantly increase production costs for most medical device
suppliers, other health-related manufacturers, or pharmaceutical
manufacturers. The very low TAC/S values for these industry
groups support this expectation. Although median values do not
indicate significant impacts, individual facilities in each of
these industries might incur significant adverse impacts.
Spice manufacturers incur relatively low unit compliance
costs, as shown by TAC/CV and TAC/EO. At the same time, they
incur some significant increases in sterilization costs, as
measured by TAC/C. Their median TAC/C is the second highest,
almost 45 percent under Option 1, 52 percent under Option 2, and
68 percent under Option 3. Also, the most severely affected
facilities in this sector incur compliance costs greater than
their baseline costs under Option 1, nearly 1.5 times their
baseline costs under Option 2, and nearly twice their baseline
costs under Option 3. Fortunately, sterilization represents a
small proportion of total production costs in this industry
(TAC/S is less than one percent, even under Option 3). Also, a
good substitute exists for EO fumigation in the spice
manufacturing industry: radiation. As described in Section 8.1,
an industry source stated that radiation has several advantages
over EO fumigation.16 These advantages include killing all
bacteria, rather than only a large percentage of bacteria.
Second, radiation can be done under ambient conditions. Third,
8-75
-------
radiation leaves no residue and requires no aeration. It does,
however, entail the costs of purchasing the equipment needed to
irradiate the product and training employees to use the new
equipment. No information is available about the relative costs
of radiation and EO as sterilization techniques. The candidate
NESHAP will probably significantly affect many contract1
sterilizers. For the other industry groups in this study,
sterilization is one of many steps in the production process. In
contrast, sterilization is nearly the entire "product" for
contract sterilizers. (As discussed in Section 8.1, many
j
contract sterilizers may also perform some packaging, testing,
and distribution services that are linked to their sterilization
operations.) Thus, the candidate NESHAP will probably cause a
more pronounced increase in contract sterilizers' production
costs, which is reflected in the relatively high median TAC/S
(2 percent under Option l, 3.9 percent under Option 2, and
4.6 percent under Option 3) for contract sterilizers. [JAs noted
earlier, one of the shortcomings of total sales is that lit
I
includes sales of sterilized and nonsterilized products;! because
virtually all sales from contract sterilizers involve sterilized
products, this shortcoming would tend to result in higher TAC/S
for these facilities than for facilities in the other iridustry
groups.] i
In addition to significantly increasing their sterilization
costs, the candidate NESHAP may result in increased demand for
I
contract sterilization services. Because contract sterilizers on
[
average have larger chambers than the other industry groups and
run more EO through them, the per-unit cost of the candidate
NESHAP is less for contract sterilizers than for the other
groups. This conclusion is supported by the data in Tables 8-14
through 8-19, which indicate that contract sterilizers have the
lowest median TAC/CV, TAC/EO, and TAC/C of the seven industry
groups under all three control options. As explained in!
Section 8.7, the contract sterilizers' lower per-unit control
costs may actually cause them to gain additional business if
i
8-76
-------
firms switch from in-house sterilization to contract
sterilization.
Aside from the effects of the candidate NESHAP controls on
the museums and libraries group, the contract sterilizers group,
and perhaps the spice manufacturers group, significant effects
may occur at some individual facilities in the other five
industry groups. In general, these are the facilities with the
highest TAC/S values in the groups. These facilities may choose
to incur the relatively high control costs, to switch to another
sterilization process, to switch from in-house to contract
sterilization, or to discontinue their production of sterilized
products. Without further information on these facilities, there
is no way to predict which response will be chosen in each case.
8.8.2 Effects on New Facilities
Up to this point the analysis has focused on the economic
effects of the candidate NESHAP controls on existing facilities.
Without data on possible control costs for new (not-yet-
constructed) facilities in each industry group, making any
quantitative estimates of the potential effects of the candidate
NESHAP, under each control option, on these facilities was. not
possible. Nevertheless, some general conclusions may be reached
on this matter based on the information presented earlier in this
chapter.
The supply curve for products sterilized using EO will shift
upward as a result of the candidate NESHAP controls. If the
demand curves for these products are at all elastic, the quantity
of the products sold will decrease. This decrease will delay
investment in new facilities that would use EO.
The effect on investment by contract sterilizers is more
difficult to predict. Because some facilities in the other six
industry groups may decide to switch from in-house sterilization
to contract sterilization, the demand curve for contract
sterilization may shift out as a result of the candidate NESHAP
controls. The market share of contract sterilization will
increase, and the absolute quantity of contract sterilization may
increase. Depending on the change in the profitability of
8-77
-------
contract sterilizers resulting from the candidate NESHAP
controls, some additional investment by contract sterilizers
might occur. i
8.9 REFERENCES |
i
1. U.S. Environmental Protection Agency. Ethylene Oxk.de
Commercial Sterilization Database. Research Triangle
Park, NC. 1987, updated 1989.
2. Telecon. Adams, P. Sterilization Services, with Midwest
Research Institute, March 2, 1987. j
3. U. S. Environmental Protection Agency. Technical Report:
Ethylene Oxide Emissions from the Use of Ethylene Oxide as a
Sterilant at Commercial Sterilization Facilities. [Research
Triangle Park, NC. 1986. i
4. Midwest Research Institute. Industry Profile: I
Sterilization (Draft). Raleigh, NC. March 10, 1987.
8
10
11.
12,
13
Federal Register.
pp. 40285-89.
1985. Vol. 50, No. 19. October 2,
Layard, P. R. G., and A. A. Walters. Microeconomics Theory.
New York, McGraw-Hill. 1978. j
1 .
Telecon. DeMarco, M. Johnson and Johnson, Inc., with
L. McNeilly, Research Triangle Institute, April 16\ 1987.
Cost of production for commercial sterilization facilities.
Telecon. Jorkasky, J. F., Health Industry Manufacturers
Association, with MRI. 1987. :
U.S. Department of Commerce, International Trade i
Administration. 1989 U.S. Industrial Outlook.
Washington, DC. 1989.
U.S. Department of Commerce, Bureau of the Census.
1987 Census of Manufactures: Industry Series, Preliminary
Report. Washington, DC. May 1989.
U.S. Department of Commerce, International Trade
Administration. 1987 U.S. Industrial Outlook.
Washington, DC. 1987. i
I
U.S. Department of Health and Human Services, Food jand Drug
Administration. Sterile Medical Devices: A GMP Workshop
Manual. Fourth edition. HHS Publication FDA 84-4174.
Washington, DC. 1984.
U.S. Department of Commerce, Bureau of Census. Census of
Manufactures: Industry Series. Washington, DC. 1987.
8-78
-------
14. Beloian, A. "Alternatives and Economic Impacts of
Cancelling Ethylene Oxide for Non-Medical Uses." Memo to
U.S. Environmental Protection Agency. June 11, 1985.
15. Weber, F. E. "Ethylene Oxide, Under Fire as a Fumigant, is
Still Most-Effective Spice Sterilant," Food Engineering.
May 1980.
16. Telecon. Purvis, J., A.C. Legg Packaging Company, with
D. L. Newton, Midwest Research Institute, March 14, 1986.
17. U.S. Department of Commerce, Bureau of the Census. Census
of Services. Washington, DC. 1985.
18. Telecon. Burns, T., American Spice Trade Association, with
A. S. Ross, Research Triangle Institute, July 24, 1987.
Data on spice industry.
19. Edwards, S. R., B. M. Bell, and M. E. King. Pest Control in
Museums: A Status Report (1980). Association of Systematic
Collections, Norman, OK. 1981.
20. Telecon. W. V., Smithsonian Museum Support Center, with C.
Beal, Midwest Research Institute, November 21, 1985. EO use
in museums.
21. Telecon. Beal, W., Methodist Church Archives and History
Center, with A. S. Ross, Research Triangle Institute,
July 17, 1987. EO use in museums.
22. Telecon. Sand, D., First Church of Christ Scientist,
Christian Science Center, with A. S. Ross, Research Triangle
Institute, July 17, 1987. EO use in museums.
23. Telecon. Hanthorn, I., Parks Library, Iowa State
University, with A. S. Ross, Research Triangle Institute,
July 22, 1987. EO use in museums.
24. Telecon. Falco, C., U.S. Department of Agriculture,
Structural Pesticides Division, with A. S. Ross, Research
Triangle Institute, July 22, 1987. Pesticide regulations
for VIKANE and EO.
25. Telecon. Colglazier, D. L., Old Sturbridge Village
Conservation Lab, with A. S. Ross, Research Triangle
Institute, July 17, 1987. Fumigation in museums.
26. Telecon. Lowery, A., U.S. Food and Drug Administration,
with C. Beal, Midwest Research Institute, October 11, 1985.
27. Rice, L. L. Memo to D. Markwordt, U.S. Environmental
Protection Agency, including a list of contract sterilizers.
April 18, 1986.
8-79
-------
28. Telecons with six contract sterilizers who each requested
that their information be treated as confidential.1
r
29. Dunford, R. W., et al. Economic Analysis of Candidate
hazardous Organics NESHAP Controls: Draft Report., Research
Triangle Institute, Research Triangle Park, NC.
November 1986. '•
30. Midwest Research Institute. Cost Effectiveness of[Reducing
Ethylene Oxide Emissions from Sterilizer Vents and j
Associated Vacuum Pump Drains. Draft Memo to D. Markwordt,
March 1990. |
i
31. Midwest Research Institute. Costing Methodology for the
Control of Aeration Room Emissions, Commercial Sterilization
NESHAP. Draft Memo to D. Markwordt, April 1990. !
32. Midwest Research Institute. Costing of Control Alternatives
for the Rear Chamber Exhaust, Commercial Sterilization
NESHAP. Draft Memo to D. Markwordt, April 1990. |'
33. Telecon. Morin, R., Plant Manager, Charles River i
Laboratories, Raleigh, NC. with A. S. Ross, Research
Triangle Institute, December 17, 1987. Cost of !
sterilization relative to total production cost. i
8-80
-------
APPENDIX A.
EVOLUTION OF THE BACKGROUND INFORMATION DOCUMENT
-------
-------
TABLE A-l. EVOLUTION OF THE BACKGROUND INFORMATION DOCUMENT
Date
October 2, 1985
October 1985
January 8, 1986
January 13, 1986
January 16, 1986
April 2, 1986
April 1986
July 1986
August 1987
August 27, 1987
September 30, 1987
November 13, 1987
March 11, 1988
March 23, 1988
Event
The EPA announces intent to list ethyl ene
oxide (EO) as a hazardous air pollutant
under Section 112 of the Clean Air Act (Act)
(50 FR 40286) .
Collection of background information begun
by Midwest Research Institute.
Site visit to Sterilization Services of
Tennessee to observe sterilization and gas
reclamation facilities.
Site visit to North Carolina Archives and
Records to observe EO fumigation chamber and
obtain data on EO use.
Site visit to McCormick and Company, Inc.,
to observe EO fumigation chambers and Deoxx™
control system.
Meeting with Johnson and Johnson (J&J)
International and Damas Corporation to
discuss the EO scrubber manufactured by
Damas and used by J&J.
Data received from a Health Industry
Manufacturers' Association (HIMA) survey
performed in November of 1985 are compiled
in the Commercial Sterilization data base.
Questionnaires sent to miscellaneous
sterilization and fumigation facilities.
Responses were received from 113 of these
facilities.
Mail out Chapters 3-5 of the background
information document (BID) for review.
Work Group briefing.
Meeting with HIMA short-term exposure limit
task force to discuss industry's response to
the EO and chlorofluorocarbon (CFC) BID
mailouts .
Second Work Group briefing.
NAPCTAC mailout of BID chapters and
Appendices.
Docket No. A- 88 -03 (Commercial
Sterilization) is submitted.
A-l
-------
TABLE A-l. (continued)
Date
May 19, 1988
July 1988
December 9, 1988
December 9, 1988
February 21, 1989
November 3, 1989
December 4-6, 1989
December 12, 1989
June 14, 1990
August 7, 1990
January 30, 1991
Event |
NAPCTAC meeting.
Questionnaires sent but to miscellaneous
sterilization and fumigation facilities.
Responses were received from 44 facilities.
Site visit to lolab, Inc., to obtain1
information about the DM3 Catcon catalytic
oxidation system used to control emissions
from the aeration room. i
i
Site visit to Medtronic, Inc., to obtain
information about the acid -impregnated
carbon adsorbtion system used to control
emissions from the aeration room.
1
Teleconference with HIMA to discuss progress
on the standard.
Meeting with HIMA to discuss progress on the
standard. i
Vendor -sponsored test of Donaldson Eto
Abater™ catalytic oxidizer.
Review of a summary of the prevalence of
chamber exhaust use among HIMA members.
Site visit to Isomedix Operations, Iric., to
obtain information about their sterilization
processes.
Work Group meeting on regulatory
alternatives . :
NAPCTAC meeting. ;
A-2
-------
APPENDIX B.
INDEX TO ENVIRONMENTAL CONSIDERATIONS
-------
-------
APPENDIX B.
INDEX TO ENVIRONMENTAL CONSIDERATIONS
This appendix consists of a reference system which is cross-
indexed with the October 21, 1974. Federal Register (39 FR 37419)
containing Agency guidelines for the preparation of Environmental
Impact Statements. This index can be used to identify sections
of the document which contain data and information germane to any
portion of the Federal Register guidelines.
B-l
-------
APPENDIX B.. i
INDEX TO ENVIRONMENTAL IMPACT CONSIDERATIONS
Agency guidelines for preparing
regulatory action environmental
impact statements (39 FR 37419)
Location within the background
information document (BID).
1. Background and description
Summary of the regulatory
alternatives
Statutory authority
Industry affected
Sources affected
Availability of control
technology
2. Regulatory alternatives
Regulatory alternative X
Environmental impacts
Costs
Regulatory alternative B
Environmental impacts
Costs
The regulatory alternatives are
summarized in Chapter 1.
Statutory authority is; given in
Chapter 1 and Chapter 2.
A description of the ihdustry
to be affected in given in
Chapter 8. I
Descriptions of the various
sources to be affected; are
given in Chapter 3. ;
Information on the availability
of control technology is given
in Chapter 4.
Environmental effects of
regulatory alternative A are
considered in Chapter 6.
Costs associated with
regulatory alternative|A are
considered in Chapter 7.
Environmental effects of
regulatory alternative jB
emission control systems are
considered in Chapter 6.
Costs associated with i
regulatory alternativejB
emission control systems are
considered in Chapter 7.
B-2
-------
Agency guidelines for preparing
regulatory action environmental
impact statements (39 FR 37419)
Location within the background
information document (BID).
Regulatory alternative C
Environmental impacts
Costs
Regulatory alternative D
Environmental impacts
Costs
Regulatory alternative E
Environmental impacts
Costs
Environmental effects of
regulatory alternative C
emission control systems are
considered in Chapter 6.
Costs associated with
regulatory alternative C
emission control systems are
considered in Chapter 7.
Environmental effects of
regulatory alternative D
emission control systems are
considered in Chapter 6.
Costs associated with
regulatory alternative D
emission control systems are
considered in Chapter 7.
Environmental effects of
regulatory alternative E
emission control systems are
considered in Chapter 6.
Costs associated with
regulatory alternative E
emission control systems are
considered in Chapter 7.
B-3
-------
-------
APPENDIX C.
EMISSION SOURCE TEST DATA
-------
-------
APPENDIX C.
EMISSION SOURCE TEST DATA
This appendix contains summaries of performance tests EPA
and industry conducted on four acid^-water scrubbers and one
catalytic oxidizer designed to control ethylene oxide (EO)
emissions from sterilizer exhaust and aeration room gas streams,
respectively. Performance tests to determine control device
efficiency were conducted on two types of acid-water scrubber
systems, Damas™ and DEOXX™. Detailed descriptions of these types
of scrubbers are presented in Chapter 4. The sterilizers tested
use pure EO and a 12/88 mixture of EO and chlorofluorocarbons
(CFC's) as sterilant gases. The results of the five tests are
presented in the following sections of this appendix:
C.I—EPA test of a DEOXX™ system at Burron Medical;
C.2—Independent laboratory test of a DEOXX™ system at
McCormick and Company, Inc.;
C.3—Independent laboratory test of a DAMAS™ system at
Johnson and Johnson, Inc.;
C.4—Independent laboratory test of a DEOXX™ system at
Chesebrough-Pond's DEOXX™ facility; and
C-5—Vendor test of an EtO ABATOR™ system at Seamless, Inc.
A summary of the test results and selected test conditions
for the first four tests is provided in Table C-l.
C.I EPA TEST OF A DEOXX™ SYSTEM AT BURRON MEDICAL
C.I.I Facility Description1
An EPA-sponsored test was conducted on a Deoxx™ acid-water
scrubber in September 1987. The test took place at Burron
Medical, a medical supply sterilization facility located in
Allentown, Pennsylvania. The facility has three 28 cubic meters
C-l
-------
W
EH
g
CO
§
^
u
w
B
fe
O
X
w
H
\
U
H
1
.___
%
& fc 1»
III
1 §
81g
•s 1 §:
||
III
fi'
* *M
ft
O jj
*p
O "3
0 |
1
M
-
.3
""«^
o ^
i
I
1
S
o
aJ
*
•°
^
2
1
wJ*
eo
cs
I
§
M
_o
1
1
eo
eo
?
00
i
•ft
V
i
*?. -
•o
«
PureEO
M>
S
3
i
§
•»w
mick & Company
8
S
as
sis;
eo «•>
8.2
22
.n .0
s-s-
*£* S
S|
° 0
>o «
00 00
5 5
•°
f
&
n Si Johnson
|
3S;
s si
A A
O 0
• —
V
i-I
•o -o
S e
£
fS"
S
_r ^
r^ t-;
I
1
x
f
I
n
s
1
i
1
i
1
,
i
i
i
i
(
i
i
i
i
i
i
e
o
Cu
1 :
_e
'•?
S
S
0
2 a
I si : :
S 5 | ,
2 *> V
I 11 :
g •»! i '
^ Is i •
a g J \
•8 fe S : '
£ > i
1 g •£ I
n ** rt '
«S o , ;
o. «> *r> i
fi . 5 -1 |
g -o o. ?
-I'll ;
S 1 ° E 1
e -o ., i
slP
«n Q ea u. '
C-2 :
-------
(TS?} (1.,OOO cubic feet [ft3]) sterilizers that use 12/88. The
gas is supplied from a common header serving all four units and
is controlled by a liquid flow meter.
A sterilization cycle typically uses 140 liters (L)
(38 gallons [gal]) of 12/88 gas mixture. On a weight basis, a
sterilization charge consumes 167 kilograms (kg) (368 pounds
[lb]) of gas, of which approximately 20 kg (44 Ib) are EO. The
initial charge of EO to the chamber was calculated using the
weight of the supply cylinders before and after charging the
chamber.
The exhaust from the sterilizers is controlled by a DEOXX™
system. At the time of the test, the scrubber contained a dilute
mixture of phosphoric and sulfuric acid. Each chamber is
equipped with a total recirculating liquid vacuum pump. These
pumps are equipped with gas/liquid separators, which emit the gas
to the DEOXX™ system and recirculate the liquid to the pump
inlet. Chambers Nos. 1 and 2 are equipped with oil-sealed pumps.
Chamber 3 is equipped with a water-sealed pump. All of the tests
were conducted using the chambers (Nos. 1 and 2) equipped with
oil-sealed pumps.
The sterilization cycle is controlled automatically by a
programmable microprocessor system. The control system can
control and record the parameters of the sterilization cycle
including chamber temperature, chamber pressure, and elapsed time
from the start of the cycle.
The sterilization process begins with a humidifying step,
which takes place in a separate room. After the humidifying
step, each load to be sterilized is transferred to the
sterilization chamber. The sterilization cycle is a batch
process that takes 4 to 6 hours. A sterilizer load begun during
the morning shift exhausts at about 2:00 p.m. In a typical plant
operating mode, seven poststerilization evacuations occur over a
3-hour period. After the chamber is repressurized, following the
seventh evacuation, the product is removed from the chamber and
allowed to off-gas. Although the control system is designed to
handle the exhaust from two sterilizers venting simultaneously,
C-3
-------
the tested sterilization cycles were scheduled so that £>nly one
sterilizer vented at a time.
Three different sterilization programs were used for
testing: one for the empty chamber tests, one for the full
chamber tests, and one for the last full chamber test (Test 13).
The testing of both full (loaded) and empty chambers was carried
out to examine the effects of product retention of EO oh the EO
emissions. A total of 17 tests were performed, 5 with product in
the chamber (full chamber tests) and 12 without product;(empty
chamber tests). Seven of these tests were considered invalid
according to the testing lab, therefore, data from 10 of these
' i
tests (4 with and 6 without product) were summarized in the final
report. Table C-2 summarizes operating data from the 10 valid
tests.
Before the start of each test (except Test 13), the chamber
was evacuated to 2 pounds per square inch absolute (psia) and
then pressurized to 3.1 psia with steam. The humidifying step
was maintained at 3.1 psia for 1 hour for the loaded chamber
tests, but the humidifying step was shortened to 5 minutes for
the empty chamber tests. After the humidifying, the chamber was
charged to 23.9 psia with 12/88 gas. The exposure at 23.9 psia
was maintained for 4 hours for the loaded chamber tests j but was
shortened to 5 minutes for the empty chamber tests. During the
last full chamber test (Test 13), the chamber was evacuated to
7 psia and pressurized to 32.9 psia. !
i
Each test program contained seven poststerilization
evacuations; an initial chamber evacuation, six air in-bleeds
with subsequent evacuations, and a final air in-bleed. |Except in
Test 13, the chamber was evacuated to 2 psia and pressurized with
air to 13.9 psia during each evacuation and air in-bleed cycle.
The initial chamber evacuation and pump down lasted 26 to
27 minutes. Each subsequent evacuation lasted 12 to 14 iminutes,
and each air in-bleed required 12 to 14 minutes. During Test 13,
the chamber was evacuated to 7.0 psia, which reduced each
evacuation and air in-bleed time by 7 minutes. ;
C-4
-------
3
oo
00
(M
o
c4
VO
o\
.1
5
S
8
013
oo
o
-3
i
o
r-
oo
§
o
o
OO
VO
VO
PI
oo
oo
fS
ON
tn
eo
IS
. t 3= S,
« s • a •§ «.
i|-Mii
fn o «n § oo *
2 § 2 | 3 e
S I S | E £
CM
I
u
w
o
ha
E
04
cs
uc
o
2
o
0
£
I
VO
eg
oo
o\
o
en
>o
lillll
&•* B.w> O.~"
BO -g H)>5 M -o
£ £ £ £ £ «
II! Hi
•5 ta '55 w -K rS
o. > a o.10
r4 oo ^ oo t~-
C-5
-------
C.I.2 Sampling Procedures1
The sampling and analysis procedure used in these field
tests involved semicontinuous direct sampling of scrubber inlet
and outlet gas streams with on-line gas chromatographic analysis
to determine EO concentrations. Volumetric flow rates both
entering and exiting the scrubber were determined with orifice
plates. However, some difficulty was encountered when measuring
the inlet volumetric flow rate due to the high EO concentration
and low flow rate of the gas stream.
C.I.3 Test Results1
Table C-3 presents uncontrolled and controlled emissions
(i.e., the amount of EO entering and exiting the control unit)
for the six empty chamber tests. Emission measurements; were
considered for the six empty chamber tests because the total
amount of EO entering the scrubber were more consistent for these
runs. These emissions measurements were used to evaluate the EO
control efficiency of the DEOXX™ system. :
Efficiencies were calculated for the six test runs!.
Throughput efficiency was calculated using the gas stream EO
concentrations and volumetric flow rates into and out of the
DEOXX™ unit. This method calculates the efficiency of the
control device, i.e., the difference in mass of EO into; and out
of the control device. However, the method used to determine
this efficiency suffered due to the inability of the orifice
plates to accurately measure the flow rate into the scrubber.
The second method of determining control efficiency, recovery
efficiency, was calculated using the weight of the original EO
charge and the measured EO emissions at the outlet of the control
unit. !
Both of the analytical methods used to determine control
efficiency were hindered by the difficulty in determining the GC
peak for EO scrubber outlet concentration measurements. This
problem arose from a shift in the EO retention time (in!the
column of the GC) as EO concentrations exiting the scrubber
decreased. The difficulty in identifying the EO peak was further
C-6
-------
U
O4
w
O
1
1
ll?
pi-
06
tr
111.
P «
s
1 o
OO bp
1 |
0
tu
_
s
o ^,
jf-*>
11
O
U
ff
.s ^
« »
O .2
U E
u
III
08 JJP
"8 .
o J
"3 g
I"8
S 2
£
2
s
a
s;
8
0
§
o
2
S
oo
0
3
o
2
o
00
«J
os
o\
o;
85
00
s
o
o
§
o
•a
bs'
?.
—
55
°
•-<
oo
o
>n
5
oo
oo
m
o\
R
.
§;
1
o
o
o
•a
cs
S
a
00
cs
cs
o
0
0
^~
cs
cs
0\
0
S;
o\
oo
OS
§;
o
o
o
8
?
OO
Os
2
0
s
o
1
M
60 60 ^
I
U
C-7
-------
complicated by the large range of EO concentrations exiting the
scrubber. ;
Statistical tests were performed to evaluate the effect of
product in the chamber on determining the efficiency of the
control unit. These tests showed that the presence of product in
the chamber had no significant effect on the efficiency;
determinations.
The absolute difference between measured emissions and
expected emissions (based on the initial EO charge to the
chamber) was greater than 40 percent for three tests and less
than 10-percent for only one test. In five of the six empty
chamber tests, the measured emission levels were higher than the
expected levels. From these observations, actual uncontrolled EO
emissions may therefore be expected to be from 50 to 150 percent
of the actual emissions.
C.2 INDEPENDENT LABORATORY TEST OF A DEOXX™ SYSTEM AT McCORMICK
AND COMPANY, INC.
C.2.1 Facility Description2 .
A DEOXX™ detoxification system was installed at the Hunt
Valley Spice Mill of McCormick and Company to control EO
1 i
emissions from sterilizers. The DEOXX™ system was testjed by an
independent laboratory the week of October 14, 1985, to evaluate
performance.
The sterilizers at the Hunt Valley Spice Mill are used to
t
process a variety of spices. Pure EO is used as the sterilant.
Each sterilizer is equipped with a total-recirculation liquid
ring vacuum pump system to evacuate the chamber and achieve the
desired levels of vacuum. At the completion of the sterilization
cycle, sterilizer gas is exhausted to the atmosphere through the
DEOXX™ system. •
A test program consisting of four tests was conducted using
Sterilizer B, which had a volume of 35.3 m3 (1,248 ft3). The
first test was used to check the equipment and instrumentation
operation. .The remaining three tests (Test Nos. 2, 3, and 4)
were used to evaluate the DEOXX™ system performance. Figure C-l
provides locations of the sampling points.
C-8
-------
u
OJ
C C-
co
EH O
CO £
W
•«-
DEOXXtm
SYSTEM
EH
CO
CO
EH
CO
D
X
EH
z
I
J
O &«
M
EH CO
2 u
EH
Z U
8S
Z EH
O U
U 2
O iJ
EH O
CO
CO
EH 0.
K
U
CU EH
S rt
W EH
EH CO
CO CO
< <
(0
o
•H
.p
(0
o
o
-p
C
•H
o
a
tr>
C
Cu
g
n3
CO
I
U
cu
CO
(X
a
w
EH
CO
C-9
-------
The sterilization cycle operating conditions for the three
i
performance tests are summarized below:
Initial evacuation conditions;
Pressure, psia (in. Hg)
Time, min
14.2 (28.9)
25
Total EO charged to chamber, kg (Ibl;
Test 1 13.2 (29.0)
Test 2 13.2 (29.0)
Test 3 12.7 (28.0)
Exposure conditions!
Pressure, psia (in. Hg)
Temperature, °C (°F)
10.4 (21.2)
43.3 (110)
Following the sterilization cycle, the chamber was evacuated to
7.0 psia (14.3 in. Hg) over a period of 12 minutes. Two
additional evacuations lasted 10 minutes each, and the air washes
required less than 1 minute each.
C.2.2 Sampling Procedures2
The sterilization chamber was kept empty during the test
cycles. This eliminated possible variations in EO emissions
during the exhaust phase due to product off-gassing without
adversely affecting the performance evaluation of the DEOXX™
system. !
For each performance test cycle, the amount of EO charged to
the sterilizer was determined by measuring the weight of| the EO
supply cylinder before and after charging EO to the sterilizer
chamber. l
i1
The weights of EO entering and leaving the DEOXX™ system
were determined for each chamber evacuation by continuously
monitoring the total volumetric gas flow rate and EO j
concentration at the inlet and the outlet of the DEOXX™ jsystem.
The volumetric gas flow rate was measured by using an orlifice
meter at each location. The gas pressure drop across the orifice
plate was monitored throughout the exhaust cycle for accurate •
measurement of gas flow rate. The EO concentrations were
C-lO
-------
measured with two gas chromatographs, one each at the inlet and
the outlet of the DEOXX™ system.
C.2.3 Test Results
The weights of EO entering and leaving the DEOXX™ system for
each evacuation, as well as the removal efficiencies associated
with each test run, are presented in Table C-4.
C.3 INDEPENDENT LABORATORY TEST OF A DAMAS™ SYSTEM AT
JOHNSON & JOHNSON, INC.3
C.3.1 Facility Description
A Damas™ scrubber is used to control EO emissions from the
sterilization operations at Johnson & Johnson's Ethicon, Inc.,
facility in Somerville, New Jersey. During the week of
August 27, 1984, a Damas™ scrubber at this facility was tested by
an independent laboratory. The sterilizer uses a 12/88 mixture
of EO/CFC as the sterilant. All concentrations reported are
based on gas chromatograph (GC) analyses.
C.3.2 Sampling Procedures
Nine analyses were performed on the scrubber outlet: six
with a scrubber flow rate of 0.031 cubic meters/second (m3/sec)
(66 cubic feet per minute [ft3/min]) (Set 1) and three with a
flow rate of 0.047 m3/sec (100 ft3/min) (Set 2). These two sets
of analyses were performed on August 29 and August 30, 1984,
respectively. Samples were collected by drawing a small amount
of the scrubber emission stream through a Teflon™-lined pump to a
mobile laboratory using a Teflon™ sample line. A collection sump
with a sampling port allowed samples to be drawn with a syringe
for injection into a GC with a flame ionization detector. These
syringe samples were taken once every 2 minutes during each
analysis. The scrubber outlet emission stream was also
continuously monitored using an infrared analyzer.
The scrubber inlet gas stream was tested on August 30, 1984,
using a scrubber flow rate of 0.047 m3/sec (100 ft3/min). Three
grab samples were collected (during one evacuation) in Tedlar™
bags and analyzed in the same manner as the scrubber outlet
samples. The EO concentration at the scrubber inlet was based on
an average concentration of these three samples.
C-ll
-------
H
O
O
H
J U
O H
pj p,
EH CO
§ .
to
CO Q
co
o
Ul
60 00
•*
11
O
W
S"
c ^
«-^
° iT
o-i
CI] g
•8
If
w •§
]4
si
2,
o
s
o
s
o\
o\
00
gj
oo
oo
o\
8
§
5
5
o\
si
vo
oo
o\
8
oo
cs
•o
1
o
•8
1
I
S
• 1
S S3
11
II
g)J
-
"S
§2
s
II
11
C-12
-------
C.3.3 Test Results
The peak outlet concentration for each analysis was used to
calculate the removal efficiency of the scrubber. The average
scrubber efficiency was 99.25 percent for Set 1 and 99.28 percent
for Set 2. The average removal efficiency for all nine runs was
99.26 percent.3 (The average removal efficiency determined by
infrared analysis of the scrubber outlet EO concentration was
99.16 percent.)
The use of concentration differences, rather than a percent
weight removal, as a basis for calculating the removal efficiency
did not significantly affect the efficiency estimate; the
efficiency still should be equal to or greater than 99.0 percent.
The use of an average inlet concentration, which was based on the
highest gas flow rate, and the peak outlet concentration to
determine the efficiency would provide a conservative estimate as
long as the outlet flow rate is less than or equal to the inlet
flow rate. However, if the outlet flow exceeds the inlet flow
(e.g., if there is dilution at the stack), then the methodology
would overestimate the efficiency. It could not be determined
from the data provided whether the outlet and inlet flow rates
were different, and there were no indications in the report that
flow rates were monitored.3
Some uncertainties exist regarding the efficiencies obtained
in this test because of the conditions under which the scrubber
was tested. First, the test was performed on a scrubber that was
using fresh scrubbing liquor (i.e., no ethylene glycol in the
scrubbing liquor). Secondly, it was unclear from the test report
whether all the runs each day (i.e., number of evacuations) were
for a single sterilization cycle. Finally, there were no
indications that gas stream flow rates were monitored during the
test.
013
-------
C.4 INDEPENDENT LABORATORY TEST OF A DEOXX™ SYSTEM AT
CHESEBROUGH-POND'S DEOXX™ FACILITY4
C.4.1 Facility Description I
A DEOXX™ system was installed on sterilizer 4 at !
Chesebrough-Pond's Sherburne, New York, plant in 1982 to control
EO emissions. Four tests were conducted on December 1982 over
three different sterilization cycles to evaluate the performance
of the system. These tests were performed using pure EO and
12/88 sterilant, and were all conducted on an empty chamber.4
This chamber had a volume of 35.7 m3 (1,260 ft3) and was capable
i
of using either pure EO or a mixture of 12/88 EO/CFC as the
sterilant gas.
C.4.2 Sampling Procedures
r a i
A total of four performance tests was conducted on the
DEOXX™ system: three using the 12/88 mixture and one using pure
EO gas. The scrubber outlet emission stream was sampled once
every minute by routing two sampling streams from the scrubber
exhaust to two separate GC's. The volumetric flow rate of the
scrubber outlet emission stream was also monitored using an
orifice plate. Volumetric flow vs EO concentration profiles were
then developed to calculate the mass of EO exiting the scrubber.
The amount of EO entering the scrubber was determined by
weighing the sterilant supply cylinders before and after use, and
then subtracting the remaining sterilant in the chamber after the
cycle. The EO concentration remaining in the chamber was
measured with a GC, and the chamber pressure was measured; the
ideal gas low was used to determine the weight of EO left in the
chamber.
C.4.3 Test Results
A summary of the emission measurements and control
I
efficiencies for these emissions tests is presented in jTable C-5.
The efficiencies were calculated based on the total weight of EO
entering and exiting the scrubber over the entire cycle. The
removal efficiencies for the three 12/88 mixtures were ;
99.0 percent, 98.7 percent, and 99.4 percent, respectively, for
C-14
-------
in
w
I
5£
M 60
g
BO *
ta •"
-
*J
•2 e so
r> S •*
S
I
ON
v
NO,
*
8
NO
8
1
8
'5
I
I
§
I
•3
•u
s
1
E
II
"1
S «
•S op
oo S
•s-S
'ig
11
oo ;Sb
** 41
§1
'« -S.
.'" .2P
|j
oo oo
11
ll
3 J2
g 3
M
C-15
-------
an average of 99.0 percent. For the pure EO mixture, the EO
removal efficiency was greater than 99.99 percent.4
C.5 VENDOR TEST OF AN ETO ABATOR™ SYSTEM AT SEAMLESS, INC.5
C.5.1 Facility Description i
A Donaldson EtO Abater™ catalytic oxidizer was installed to
i
control aeration room emissions at Seamless' Ocala, Floirida,
facility. Testing was performed on the catalytic oxidizer on
December 4 through 6, 1989. The control system was installed to
demonstrate to the State that Seamless could comply with the
State's 1 ppm standard at 50 feet from the fenceline.6 This
facility sterilizes products in any one of three sterilizer
chambers using pure EO or 12/88 (EO/freon) sterilant, depending
on which sterilizer is used. After the product is sterilized, it
is taken to the aeration room, where it is then allowed!to off-
gas for at least 24 hours. |
The aeration room is maintained at a temperature of 38°C
(100°F), and the EO emissions are controlled by two 56 m3/min
(2,000 ft3/min) catalytic oxidizers. These units run
continuously and provide enough heated, recirculated air to
maintain the aeration room at a temperature of 38°C (100°F).
C.5.2 Sampling Procedures
Velocity measurements were performed using Pitot tubes
situated at 12 traverse points along the cross-section of the
duct both upstream and downstream of the control device.
Emission stream temperature measurements were also made at these
locations. The averages for each set of measurements were used
to determine the emission stream flow rates to and from the
control device, respectively.
Six 1-liter grab samples were taken simultaneously|both
upstream and downstream of the control device via test jj>orts.
Two additional 4-liter grab samples (taken in 10-liter Tedlar™
bags) were simultaneously taken both upstream and downstream of
the control device. Of the eight sets of samples, five were
taken approximately 3 hours prior to introducing sterilized
product into the aeration room, two were taken just prior to
introducing sterilized product into the aeration room, 4nd one
i
C-16
-------
was taken one-half hour after sterilized product was introduced
to the aeration room. These samples were analyzed within a
20-hour period of obtaining the samples using a gas chromatograph
with a flame ionization detector (FID). All of the product
introduced to the aeration room was previously sterilized with
pure EO.
C.5.3 Test Results
Six of the eight sets of grab samples were used to determine
the control device efficiency. From the results of the GC
analysis provided in the test report the efficiency of the
control device was determined to be 99.9+ percent for each of the
tests performed. However, these efficiencies are based on EO
concentration only and do not reflect the mass of EO entering and
exiting the control device. Also, the supporting information
provided with the test report is very limited and does not
substantiate the 99.9 percent claimed.
Other uncertainties exist in this test report. In all of
the downstream grab samples, the EO concentration was determined
to be zero (instead of the lower detection limit of the FID).
Also, the control inlet EO concentrations are high (10 ppm
minimum) for aeration room emission concentrations which
typically tend to have an EO concentration of less than 2 ppm.7
C.6 REFERENCES
1. Sampling/Analytical Method Evaluation for Ethylene Oxide
Emission and Control Unit Efficiency Determinations. Final
Report. Radian Corporation, Research Triangle Park, NC.
April 5, 1988.
2. Performance Testing Report: DEOXX™ Ethylene Oxide
Detoxification System, McCormick & Company, Hunt Valley,
Maryland, Plant. . Chemrox, Inc., Bridgeport, CT. October 29,
1985.
3. Evaluation of the Efficiency of an Ethylene Oxide Scrubber.
Scott Environmental Services, Plumsteadville, PA. October 2,
1984.
4. Certification Testing Report: Ethylene Oxide Detoxification
System on Sterilizer 4, Chesebrough-Pond's, Inc., Sherburne,
New York, Plant. Buonicore-Cashman Associates, Inc.,
Bridgeport, CT. January 13, 1983.
C-17
-------
5. Field Test Report for Seamless, Inc. Donaldson Co., Inc.
Minneapolis, MN. January 3, 1990.
i
6. Telecon. Friedman, B., MRI, with L. Cutright. Professional
Medical Products Seamless Division. January 13 andj19, 1990.
Discussion about the factors affecting the probable ethylene
oxide levels in the aeration room.
7. Memorandum. deOlloqui, V., MRI, to Commercial sterilization
Files. Responses to the July 1988 Section 114 Letter.
August 22, 1990.
C-18
-------
APPENDIX D.
EMISSION MEASUREMENT AND CONTINUOUS MONITORING
-------
-------
APPENDIX D.
EMISSION MEASUREMENT AND CONTINUOUS MONITORING
D.I METHODS FOR DETERMINING ETHYLENE OXIDE EMISSIONS FROM
ETHYLENE OXIDE STERILIZERS EQUIPPED WITH CONTROL DEVICES
Since the early 1980's, concern about the toxicity of
ethylene oxide (EO) has spurred the development of methods to
accurately measure emissions from sterilizing units that use EO
as a sterilant. The U. S. Environmental Protection Agency (EPA),
the California Air Resources Board (CARB), manufacturers of EO
control devices, and EO sterilizer operators have independently
moved toward developing such a method by conducting tests of
emission control technologies. The methodologies used in several
of these tests are summarized in the following sections. Test
methods used to evaluate dilute acid hydrolytic EO scrubbing
units are discussed in Reports 1 through 6, and test methods used
to evaluate catalytic oxidation units are discussed in Reports 7
through 9. Report 10 is a summary of CARB Method 431.
The reports summarized below were generated by control
device vendors gathering data to support efficiency claims,
purchasers of control devices to either verify the manufacturer's
claims or to comply with State regulations, or by EPA in support
of method and standard development. These reports are referenced
fully in Section D.4.
D.I.I Report 1
In this test effort,1 dilute acid hydrolytic scrubber
efficiencies were determined using (1) calculated values for EO
emissions vented to the scrubber (inlet) and (2) measured values
for EO emissions exhausted from scrubber (outlet). Sampling was
D-l
-------
performed over the entire evacuation cycle, which included the
i
initial evacuation and four air washes. |
D.I.1.1 Determination of EO Mass Vented to the Scrubber.
The mass of EO vented from the sterilizer to the scrubber was
calculated by. subtracting the residual mass of EO left in the
chamber from the mass of EO charged to the chamber. The mass of
EO charged to the chamber was determined by weighing the charging
cylinder prior to and after chamber charging. Residual; EO
concentrations were measured after the sterilization cycle was
complete using the following procedure: a diaphragm pump was
used to remove a slipstream of gas through a heated Teflon™ line,
which was analyzed using a gas chromatograph (GC) equipped with a
thermal conductivity detector (TCD). The residual EO mass was
calculated based on the chamber volume, temperature, pressure,
and the residual EO concentration.
D.I.1.2 Determination of EO Mass Emitted from the: Scrubber.
The mass of EO emitted from the scrubber was calculated from
repeated measurement of the EO concentration and volumetric flow
rate found in the scrubber exhaust. Ethylene oxide
concentrations were determined by removing a slipstream! of
exhaust gas through Teflon™ tubing using a diaphragm pump. A
sample of the slipstream was analyzed once every 3 minutes
throughout the sterilization cycle using a GC equipped with a
flame ionization detector (FID).
The volumetric flow rate of the exhaust gas was measured
once each minute throughout the sterilization cycle using an
orifice meter installed in the exhaust stack. The gas flow rate
.changed continuously during the exhaust cycles.
Ethylene oxide concentrations and volumetric flow rate data
were then plotted for the initial exhaust and subsequent air wash
cycles. Mass emissions of EO were calculated for each exhaust
cycle and totalled. Control device efficiency was determined
based on the calculated total emissions to the scrubberj and
measured total emissions from the scrubber. i
D.I.1.3 Results. The results of three scrubber efficiency
tests performed while a 12/88 mixture of EO and
D-2
-------
dichlorodi-fluoromethane was used as a sterilant showed an
average EO removal efficiency of 99.1 percent. Removal
efficiencies ranged from 98.7 to 99.4 percent. More than
78 percent of the total EO was emitted during the first exhaust
cycle.
A single scrubber efficiency test was performed using
100 percent EO as a sterilant. This test demonstrated a removal
efficiency of 99.998 percent. Due to technical problems, data
for the initial evacuation and the first air wash cycles were not
available; thus, the reported amount of 0.00105 pound of EO
exhausted was determined from measurements of subsequent air
washes.
D.I.2 Report 2
D.I.2.1 Methodology. Report 2 discusses the test methods
and results of three efficiency test runs performed on a dilute
acid hydrolytic scrubber.2 During testing, the sterilizer
chamber was empty of product, and 100 percent EO was used as the
sterilant. The weights of EO entering and leaving the scrubber
were determined by continuously measuring the total volumetric
gas flow rate at both the inlet and outlet of the scrubber with
orifice meters. Sampling was performed over the entire
evacuation cycle, which included the initial evacuation and two
air wash cycles. The inlet EO mass was calculated using the
difference by weight of the EO supply cylinders and residual EO
left in the chamber (see Section D.I.I). However, because it was
unclear exactly how residual chamber concentrations were
determined, the calculation method will not be discussed.
D.I.2.2 Determination of EO Mass at the Inlet and Outlet of
the Scrubber. Samples were withdrawn continuously from the two
locations through heated Teflon™ lines using Teflon™-lined pumps.
Slipstreams of gas were sampled with gas sampling valves at
approximately l-minute intervals into two GC's. A GC/TCD was
used to measure the inlet slipstream and a GC/FID was used to
measure the outlet slipstream.
Volumetric flow rate measurements were performed at both
sampling locations. Two orifice meters of different sizes were
D-3
-------
used at each location to supply flow measurements over the range
of expected velocities. ;
D.I.2.3 Results. The three scrubber performance tests
yielded an average removal efficiency of 99.988 percent by
weight, with individual values ranging from 99.986 to
99.989 percent. The EO in the initial evacuation accounted for
20 percent or less of the total mass of EO emitted during the
sterilization cycle.
D.I.3 Report 3
D.I.3.1 Methodology. In this test program, dilute acid
hydrolytic scrubber efficiencies were determined using calculated
and measured inlet values for the EO emissions vented to the
scrubber and measured values for the EO emissions exhausted from
the scrubber.3 Sampling was performed over the entire evacuation
cycle, which consisted of the initial evacuation and six air
washes. This report compared the "throughput" and "recovery"
methods of calculating dilute acid hydrolytic scrubber efficiency
(see Section D.I.3.3). Data from 10 test runs were repbrted.
Four test runs were conducted with an empty chamber. All runs
used 12/88 as the sterilant. ;
D.I.3.2 Determination of EO Mass at the Inlet and Outlet of
i
the scrubber. The concentration of EO entering and leaving the
scrubber was measured semicontinuously at the inlet and outlet
with a GC/FID. Sample gas was continuously removed from the
sampling locations and analyzed at 4-minute intervals. The
volumetric flow rate at the outlet of the control device was
measured by a vane anemometer in series with orifice plates; the
flow rate at the inlet was calculated as discussed below.
D.I.3.3 Efficiency Determinations. The throughput
efficiency was calculated using measured EO emissions ftom the
inlet and outlet of the scrubber. Ethylene oxide concentration
was measured at the inlet by GC/FID, and inlet volumetric flow
rates were calculated using the chamber volume and chamber
pressures and temperatures. ;
The recovery method calculated EO control efficiency based
on the weight of EO charged to the sterilization chamber. The EO
i
D-4 i
-------
charge cylinder was weighed prior to and after charging the
chamber, and the weight of EO charged was determined by the
difference. The measured outlet EO emissions were used to
calculate the recovery efficiency.
D.I.3.4 Results. Only the results for the empty chamber
tests will be discussed since tests where the sterilization
chamber was loaded with product showed similar removal
efficiencies. Throughput efficiencies for the empty chamber
tests ranged from 99.82 to 99.98 percent. Recovery efficiencies
for the empty chamber tests ranged from 99.90 to 99.97 percent.
A one-way analysis of variance (ANOVA) performed on data from the
empty chamber tests showed no difference between the recovery
method and the throughput method in determining control
efficiency.
D.I.4 Report 4
D.I.4.1 Methodology. This report focused on several issues
raised by the testing described in Report 3.4 The first issue
concerned the stability of EO samples in 1-liter (L)
(0.264 gallons [gal]) polyvinylfluoride gas bags, 5-milliliter
(mL) (0.3 cubic inches [in3]) gas-tight syringes, and evacuated
aerosol cans. The second issue investigated was EO concentration
profiles plotted at 1-minute intervals throughout the
sterilization cycle using both pure EO and 12/88 mixtures. The
third issue was the evaluation of various analytic columns.
The sampling and analytical methods reported varied slightly
from those in Report 3. Scrubber exhaust samples were removed
from the exhaust stack through a heated Teflon™ line with a
diaphragm pump. Samples were placed into 1-L polyvinyl fluoride
bags, 5-mL gas-tight syringes, or evacuated aerosol cans.
The linearity, efficiency, resolution, retention time,
sample stability, and limits of quantitation and detection were
evaluated for three columns.
D.I.4.2 Results. Report 4 concludes that a 5 percent
fluorinated oil column is adequate for measuring percent levels
of EO and dichlorodifluoromethane, the levels expected at the
exhaust of an uncontrolled sterilizer or the inlet of a scrubber.
D-5
-------
A 1 percent polyethylene glycol and substituted terephthalic acid
column is recommended for measuring par t-per-million-by-r volume
(ppmv) levels of EO, the levels expected at the outlet of a
scrubber. The report recommends that a 3-percent polyethylene
glycol and substituted terephthalic acid column operated at 45°C
(116°F) be investigated for quantifying sub-ppm levels of EO, the
levels expected in ambient air at sterilization facilities.
The stability of 12/88 and 100 percent EO samples in Tedlar™
bags were investigated by measurement at time zero and selected
intervals thereafter. Both types of samples were not stable over
a 4-day period; concentrations of the 12/88 mixture differed by
23 to 96 percent, while the concentrations of the 100 percent EO
differed by 15 to 30 percent. Only the 12/88 samples were
analyzed at a shorter time interval. After 1 day, the 12/88
mixture concentrations differed 3 to 27 percent from those at
time zero, which indicated poor stability.
Aerosol cans containing 100 percent EO were analyzed at time
zero and 4 days later. Over this period the EO concentrations
differed an average of 15 percent (range +2.82 to -44.6).
Syringe samples were analyzed at time zero and after 4 days and
were also not found to be stable. The percent difference in
concentration for both the 100 percent EO and 12/88 mixtures
averaged 23 percent.
Concentration profiles showed that EO concentrations in
sterilizer exhaust increased linearly between 5 and 13 minutes
after the start of the cycle, reached a plateau between 13 and
20 minutes, and dropped off after 20 minutes.
D.I.5 Report 5
D.I.5.1 Methodology. Report 5 addresses six test tuns
conducted at a scrubber exhaust flow rate of 1.8 cubic meters per
minute (m3/min) (66 cubic feet per minute [ft3/min]) and three
conducted at a scrubber exhaust flow rate of 2.8 m3/min
(100 ft3/min).5 Flow through the scrubber appears to have been
set for this test program, as no record of volumetric flow rate
measurement is present in the report. A 12/88 EO mixture was
D-6
-------
used as the sterilant. The number of evacuations tested in each
sterilization cycle was not reported.
Scrubber inlet samples were acquired three times during the
entire test effort. Grab samples were taken using polyvinyl-
fluoride bags and analyzed in GC/FID, as described below. The EO
concentrations of the three grab samples were 15.0, 27.4, and
29.0 percent. The average concentration of the three samples
(23.8 percent) was used as the scrubber inlet concentration. The
removal efficiencies were calculated using the average inlet
concentration and the peak outlet concentrations.
The scrubber outlet emissions were sampled using a heated
Teflon™ line and a Teflon™-coated pump. Grab samples were
removed from the sample line at approximately 2-minute intervals
using a syringe. The remaining sample passed into an infrared
(IR) spectrometer. The syringe samples were analyzed by a
standardized GC/FID. The IR spectrometer used to provide a
second measure of scrubber outlet EO concentrations was set at a
wavelength of 3.3 microns to reduce interferences.
D.I.5.2 Results. All of the efficiency results were based
on the average measured inlet concentration of 23.8 percent EO
and the maximum outlet EO concentration measured for a particular
sterilization cycle. The variation in EO inlet concentrations
indicates poor precision in determining scrubber inlet EO
concentrations by this method. The efficiency determined by
using the GC averaged 99.26 percent (99.16 to 99.32 percent
range). The average efficiency determined by using IR
spectroscopy was 99.16 percent (99.03 to 99.21 percent range).
D.I.6 Report 6
D.I.6.1 Methodology. In Report 6, EO removal efficiencies
for a dilute acid hydrolytic scrubber were calculated by
measuring EO concentrations at the scrubber inlet and outlet,
scrubber outlet volumetric flow rates, and scrubber inlet and
outlet temperatures.6 Testing was performed under laboratory
conditions on a full-scale acid-hydrolysis system. Each test
cycle included two evacuations and use of a 12/88 mixture of
sterilant.
D-7
-------
Inlet and outlet EO concentrations were measured by GC/FID.
The sample gas was removed from both ducts via Teflon™ tubing.
Grab samples were removed from the tubing with a syringe every
40 seconds. Volumetric flow rate was measured at the scrubber
|;
outlet location using a calibrated dry test meter. Scrubber
inlet flow rate was calculated by adding the EO volumetric flow
rate. Using the measurements of EO concentration, temperature,
and flow rate, the EO removal efficiency was calculated based on
the total amount of EO that entered and left the scrubber system.
Four runs were performed on actual sterilizer exhaust, and
three were performed with simulated sterilizer exhaust. For the
i
simulated exhaust, a 12/88 mixture of sterilant gas was injected
into the ductwork leading to the scrubber.
D.I.6.2 Results. Destruction efficiencies for the runs
with actual exhaust ranged from 99.995 to 99.998 percent. The
destruction efficiency for the EO injection runs averaged
99.999 percent. It was noted that of the EO charged to!the
sterilizer chamber, 25 to 56 percent was exhausted to the
scrubber during the initial evacuation. A more accurate inlet
flow rate determination would provide more representative
efficiency results.
D.I.7 Report 7
D.I.7.1 Methodology. Report 7 describes the method used to
determine the EO removal efficiency of a catalytic oxidation unit
controlling aeration room emissions.7 The sterilant was
100 percent EO, and six test runs were performed. Presilrvey
testing was conducted where a 12/88 mixture was/ used and four
runs were performed. i
D.I.7.2 Determination of EO Mass at the Inlet and'Outlet of
;bhe Scrubber. The volumetric flow rates of the inlet and outlet
sample gas streams was measured by traversing the ducts;with a
standard pitot tube according to EPA Methods 1 and 2 (40 CFR
Part 60, Appendix A). Gas temperature and relative humidity were
measured at the same locations. •
Inlet and outlet grab samples were collected simultaneously
in polyvinylfluoride gas bags using heated Teflon™ sampling lines
D-8
-------
and Teflori**-coated pumps. The bag samples were analyzed by
GC/FID within 20 hours of sample acquisition.
D.I.7.3 Results. The concentrations at the control device
outlet were below the detection limit of the GC, and EO removal
efficiencies were assumed to be greater than 99.9 percent. The
EO removal efficiencies determined using presurvey samples ranged
from 99.4 to 99.6 percent. The report speculated that the
variation in efficiencies was due to the type of sterilant used
and differences in chamber operating parameters.
D.I.8 Report 8
D.I.8.1 Methodology. Report 8 presents the results of
seven efficiency test runs performed on a catalytic oxidation
system used to control EO emissions from a sterilizer.6 Three
test runs were performed on actual sterilizer exhaust and four
test runs were conducted while EO cylinder gas was injected into
the ductwork leading to the catalytic oxidation system inlet.
Each test cycle included two evacuations, and the sterilant was a
12/88 mixture.
The gas stream entering the catalytic oxidation system was
continuously monitored for EO using a total hydrocarbon analyzer
equipped with an FID. Additional inlet EO concentration
measurements were made on grab samples using GC. An EO
concentration profile for the catalytic oxidizer outlet was
derived from GC analysis of grab samples. The sample gas streams
were removed at the catalytic oxidizer inlet and exhaust and
sampled using Teflon™ tubing and a pump. Grab samples were
removed from this tubing with gas-tight syringes at 40-second
intervals during the exhaust cycle.
Flow rate measurements were performed at the catalytic
oxidizer inlet using a standard Pitot tube. A traverse of the
duct was performed, and the volumetric flow rate was measured
subsequently at a point of average velocity of the traverse. .The
outlet flow rate was assumed to be equal to the inlet flow rate.
This sampling point was located after ambient air was added to
the unit. Although the flow rate exhausted from the sterilizer
diminished over each evacuation cycle, the flow at the catalytic
D-9
-------
oxidizer test point did not change significantly, in comparison
to dilute acid hydrolytic scrubbers, catalytic oxidizers exhibit
relatively stable inlet and outlet flow rates because aiiibient air
is added prior to the unit. Catalytic oxidizer inlet and outlet
gas temperatures were also monitored.
Using the measurements of EO concentration (inlet and
outlet), temperature, and flow rate, the EO removal efficiency
was calculated based upon the total mass of EO that entered the
catalytic oxidizer and the total mass of EO that left the unit.
D.I.8.2 Results. The EO removal efficiencies forithe
sterilizer discharge tests ranged from 99.16 to 99.40 percent.
Removal efficiencies for the duct injection tests ranged from
99.89 to 99.98 percent !
D.I.9 Report 9
D.I.9.1 Methodology. Report 9 describes the methods used
to determine the EO removal efficiency of a catalytic oxidation
unit used to control aeration room exhaust.8 Since EO
concentrations to the catalytic oxidizer were approximately
2 ppmv under normal operating conditions, pure EO cylinder gas
was added at the inlet duct to yield EO concentrations of
100 ppmv. Gas samples were acquired at the inlet to the control
device, after each of the three catalyst beds, and at the exhaust
stack. A volumetric flow rate was reported for the exhaust
stack, but no mention was made regarding the methods used to
acquire these data.
Samples were collected using hydrogen bromide (HBr)'-coated
charcoal tubes over a period of 24 hours at each of the ^five
locations described above. Each sampling train consisted of a
Teflon™ tube connected to two or more charcoal tubes in series,
connected to a sampling pump. Inlet sampling was conducted for a
20-minute period every 2 hours for the entire 24-hour test run.
Samples were recovered from the charcoal tubes in a manner
similar to NIOSH Method 1614, and the analysis was performed by
GC/FID.
i
D.I.9.2 Results. The annual EO emissions to the atmosphere
were calculated to be <0.043 kilograms per year (kg/yr)
D-10
-------
(<0.095 pounds per year [lb/yr]), based on a volumetric flow rate
of 19 actual m3/min (694 actual cubic feet per minute
[aft3/min]). The EO removal efficiency of the catalytic oxidizer
was calculated to be >99.99 percent.
As the methodology for determining volumetric flow rate was
not presented in this report, this discussion is limited to the
sampling and analytical methodology, specifically, the use of
HBR-coated charcoal tubes in conjunction with an integrated
sampling rate.
Direct GC/FID analysis of EO gas samples (Reports 1 through
9) yields a detection limit of approximately 1 ppmv. Use of the
HBR-coated charcoal tubes allows concentration of the EO sample
and a consequent decrease in the detection limit for EO in the
gas stream to 3 parts per billion by volume (ppbv). This permits
more accurate quantitation of EO concentrations in the diluted
exhaust gas streams from catalytic oxidation emission control
devices.
D.I.10 Report 10
Report 10 is not a test report but rather a summary of CARB
Method 431, "Determination of Ethylene Oxide Emissions From
Stationary Sources."9 Method 431 was based on the same
methodologies described in Reports 3 and 4 (References 1 and 2),
with the exception that turbine (vane anemometer) or Roots-type
flow meters rather than orifice meters are used to determine
volumetric flow rate. Emission testing is performed on
sterilizers containing normal product loads. Volumetric flow
rate and EO concentrations of the vent gas are measured
repeatedly for the duration of the sterilization cycle. Total
emissions are calculated from curves representing flow and
concentration versus time.
As mentioned above, the CARB method requires the use of vane
anemometers or Roots-type meters to measure volumetric flow rate;
these are certified to 1.5 percent accuracy by the manufacturer.
Two or more meters may be installed in parallel if necessary to
achieve this requirement over ;the entire expected range of flow
rates. A valve is used to switch between the two meters as flow
D-ll
-------
rates change. Also required are measurements from temperature
and pressure sensors to convert the measured flow rate to
standard temperature and pressure conditions.
The sample gas is continuously removed from the exhaust
stack via heated fluoroethylene or polytetrafluoride sample line
at a rate in excess of l L/min (0.036 ft3/min). A slipstream is
removed from this line and directed to the GC sampling loop via a
gas sampling valve. The sample is analyzed as frequently as
possible (1-minute intervals for 100 percent EO and 3- to
4-minute intervals for EO mixtures). Sample loop pressure,
sample flow rate, and slipstream flow rate must be
measured/recorded during this process. Excess sample is bubbled
through a sulfuric acid solution prior to discharge. A|GC is
required for the analysis.
Also suggested are calibration gas concentrations for both
percent and ppm level analysis. Some lateral freedom is given
with respect to calibration gas makeup and concentration. It is
required that a similar midrange audit standard be usedito verify
calibration gas composition and GC performance. Audit 4"tandards
and calibration gases must be supplied and certified by separate
suppliers.
Additional guidelines for a pretest site survey, GG
preparation, flow metering, sample train setup and operation,
data reduction, and integration of the mass flow rate curve are
also provided. !
D.2 MONITORING SYSTEMS AND DEVICES
The following parameters may be monitored to indicate proper
control device operation. All monitoring equipment should be
installed, calibrated, maintained, and operated according to the
manufacturer's specifications.
The following parameters indicate proper control device
operation for counter-current packed scrubber and
reaction/detoxification towers and thus will be monitored:
1. The EO charged to the sterilizer by weighing cylinders
or by monitoring liquid flow rate using a rotameter or orifice;
and
D-12
-------
2. The ethylene glycol concentration in the scrubber liquor
using liquid level indicators or specific gravity detectors in
the tank. If the ethylene glycol concentration exceeds 60 weight
percent, the scrubber liquid must be changed.
The following parameters indicate proper control device
operation for catalytic oxidation and should be monitored:
1. The gas temperature both upstream and downstream of the
catalyst bed using a device that continuously measures these
temperatures while the control device is in operation; and
2. The amount of diluent air using a Pitot tube or other
flow measurement device.
Proper operation of a flare is indicated by the continuous
presence of a flame. Therefore, a heat-sensing device, such as
an ultra-violet beam sensor or thermocouple, shall be installed
at the pilot light to indicate the continuous presence of a
flame.
D.3 PERFORMANCE TEST METHODS
D.3.1 Test Method Background
The EPA EO test method will reference the EPA Method 2
series and EPA Method 18 (40 CFR 60, Appendix A) as its base.
Methods 2, 2A, 2C, and 2D apply for measuring flow rates from
control device exhaust. The particular method applied depends on
the size of the duct. If orifice meters (or a similar device)
are used, it may be necessary to install more than one size in
series to measure over flow rate variations.
Method 18 applies for measuring EO concentrations entering
and exiting both catalytic oxidizers and dilute acid hydrolytic
scrubbers. Using appropriate chromatographic columns and
temperature programming eliminates interferences from
dichlorodifluoromethane and potential EO degradation products.
D.3.2 Test Method Advantages/Disadvantages
Listed below is a summary of the different methodologies
that have been addressed in D.I and the advantages and
disadvantages of each technique. These factors have been
considered in developing the EPA test method.
D-13
-------
D.3.2.1 Calculation of Inlet Mass. To determine jthe inlet
mass of EO emissions entering the control device by calculation,
the amount of sterilant entering the control device must be
measured by weighing the cylinder before and after charging the
sterilizer. The residual EO left in the chamber can be
determined using the ideal gas law and would be subtracted from
the cylinder weight to calculate the inlet mass.
The principal advantage of this technique is that it avoids
the hazards of handling high levels of EO. The principal
disadvantage is that a system leak will diminish the accuracy.
However, a leak is unlikely since sterilizer sources must also
comply with health and safety rules requiring continuous
monitoring of worker exposure to EO. '
D.3.2.2 Measuring the Inlet Mass or Outlet Mass. 'Direct
measurement of the outlet mass emission rates requires
determination of the flow rate and EO concentration. Direct
measurement of the inlet mass emission rate would require an
additional set of equipment and would have the additional hazards
of handling high levels of EO. However, direct measurement would
eliminate the potential for bias created by a leak in the system.
The EO concentration can be determined by semicontinuous
sampling with a GC. However, the elution time for EO mixtures
such as 12/88 is 3 minutes, which is not frequent enough to
define the emission profile of a sterilizer. One-minute analyses
are possible for pure EO. Grab samples may be taken at|any
interval and later analyzed by GC to better define the emission
profile. However, the stability of EO in bags, syringes, and
vacu-samplers needs additional investigation.
Removing the sample gas through a heated, inert sampling
line and analyzing it immediately eliminates the potential for
sample degradation and condensation.
An orifice meter is capable of providing continuous
volumetric flow rate information. It is appropriate for
measuring volumetric flow rate in a system where flow rejtes vary
rapidly. Multiple orifice meters can be used to cover a| wide
range of flow rates.
D-14
-------
The presence of other components in the vent gas stream
creates two problems. First, components eluting near the EO peak
may create confusion in identifying and guantitating the EO peak.
Second, components that elute after EO and dichlorodifluoro-
methane extend the analysis time and decrease the number of on-
line samples that can be analyzed.
D.3.2.3 Efficiency Determinations. An efficiency
determination based on the initial evacuation of a sterilizer
does not necessarily evaluate the scrubber efficiency during
subsequent air washes. Integrating the EO concentration curve
instead of using a peak concentration value is more
representative of overall efficiency; using the peak
concentration will negatively bias the results.
D.4 REFERENCES
1.
2.
3.
4.
5.
6.
7.
Buonicore-Cashman Asociates, Inc., Certification Testing
Report; Ethylene Oxide Detoxification System on Sterilizer
No. 4 Chesebrough-Pond's, Inc. Sherburne, New York Plant.
Submitted to Chesebrough-Pond's, Inc., Corporate
Engineering. Clinton, CT. January 13, 1983.
Chemrox, Inc. Performance Testing Report: DEOXX Ethylene
Oxide Detoxification System. Submitted to [confidential
client]. Bridgeport, CT. October 29, 1985.
Radian Corporation. Sampling/Analytical Method Evaluation
for Ethylene Oxide Emission and Control Unit Efficiency
Determinations: Final Report. April 5, 1988.
Radian Corporation. Analytical Method Evaluation for
Measuring Ethylene Oxide Emissions from Commercial Dilute-
Acid Hydrolytic Control Units: Final Report. September 30,
1988.
Scott Environmental Services. Evaluation of the efficiency
of an Ethylene Oxide Scrubber. Prepared for Ethicon, Inc.,
Plumsteadville, PA.
Midwest Research Institute. Ethylene Oxide Control
Technology Development for Hospital Sterilizers. Prepared
for Atmospheric Research and Exposure Assessment Laboratory
March 2, 1988.
Klock, N., (Donaldson Co., Inc.). Field Test Report for
Seamless, Inc. Prepared for U. s. Environmental Protection
Agency. Research Triangle Park, NC. January 3, 1990.
D-15
-------
8. Engineering Science. Report of Air Pollution Source Testing
conducted at IOLAB Corporation. December 7, 1988.
9. California Air Resources Board Source Test Method 431,
Determination of Ethylene Oxide Emissions From Stationary
Sources. i
D-16
-------
APPENDIX E.
SUPPLEMENTAL INFORMATION TO THE COST ANALYSIS
-------
-------
APPENDIX E. -
SUPPLEMENTAL INFORMATION TO THE COST ANALYSIS
This Appendix contains supplemental information used in
analyzing costs associated with the regulation of ethylene oxide
(EO) commercial sterilization facilities. Included in this
appendix are: (1) costs for acid-water scrubbers (Section E.I);
(2) sample calculations of the equations used to develop capital
and annual costs for acid-water scrubbers (Section E.2);
(3) aeration room cost analysis (Section E.3); (4) capital and
annual control costs for the sterilizer chamber, chamber exhaust,
and aeration room vent(s) at an example facility (Section E.4);
(5) a breakdown of manifolding costs for these three vents
(Section E.5); and (6) the cost indices and conversion factors
used to convert costs to fourth quarter 1987 dollars
(Section E.6).
E-l
-------
E.I COSTS FOR ACID-WATER SCRUBBERS
-------
CO
•
o
CO
J ~.
S w
1
w
PL)
lJ
1
!**
*n C
I
1-
I
•*s
•rt O
8*
'w C
M
§ E
W-8
1
1
8
8
o
8
V
V
A
8
A
A
o
NO
ID 3
8
_c
•i
•§
1|
II
n
* -a fr
J 5 §.«
•S I g 1
§.2 ^* ^
Svo O *
iht*
£
-------
TABLE E-2.
Item
CAPITAL AND ANNUALIZED COSTS OF INSTALLING SCRUBBERS1
f4th Quarter 1984 Dollars]
Installed equipment costs
Automated scrubber
Explosion-proof valves for scrubber
Chlorine filter house
Purchased equipment costs, total
Installation of scrubber
Installation of chlorine filters
Taxes: 5 percent of equipment cost
Freight: 5 percent of equipment cost !•
Vacuum pumpc j
Manifolding of chambers (includes check valve)
Total capital costs, 1984 dollars
Direct operating costs
Labord
Materials
50 percent H2SO4
50 percent NaOH
Chlorine filters
Taxes: 5 percent of materials cost
Freight: 5 percent of materials cost
Compressed air
Disposal of ethylene glycol
Indirect operating costs
Overhead: 0.80 x labor
Property tax, insurance, and administration®
Capital recovery costsc
Total annualized costs
Reduce, Mg EO yr
Cost effectiveness, 1984, $/Mg EO
?Rounded to three significant figures.
£Not applicable.
^Four vacuum pumps at $4,935 each.
3.56
.
was calculated for °-25 Person-hours/shift, 3 shifts/d,
365 d/yr for system inspection and 16 person-hours for;each
regeneration of the scrubber at $11.60/person-hour.
Calculated as 4 percent of total capital cost.
E-4
-------
TABLE E-3.
DATA USED TO CALCULATE SCRUBBER EQUIPMENT
CAPITAL COSTS1
(4th Quarter 1987 Dollars)
Item
Automated scrubber
Explosion-proof valves for
scrubber
Chlorine filter house
Scrubber installation
Chlorine filter installation
Taxes
Freight
Vacuum pump(s)
Manifolding of chambers
Check valve
Cost factor
a
a b
($41.50 each) x (No. of
tanks)0
50 percent of scrubber cost
($20.00) x (No. of tanks)0
5 percent of total
equipment cost
5 percent of total
equipment cost
$5,170 per pump
— d
e
aFunction of chamber size (see Section E.I).
bExplosion-proof valves are necessary if the chamber that is
vented to the scrubber uses a gas mixture greater than
20 percent (by weight) EO. (See Section E.I).
°Number of scrubber tanks required = scrubber conversion capacity
divided by the conversion capacity of one tank (2,000 pounds of
HEO) .
aSee Section E.5.
eSee Table E-4.
E-5
-------
Cost item
TABLE E-4. CAPITAL COST OF CHECK VALVE FOR CHAMBER
~"*^^^•*—^•^^^^^^^^^^Si^^S5^^^^^^^^SS^^^^^^^^^^^^^™™™""^^^^^^^^""^^^^i^^jg^^^^^»..^
Cost
1987
Richardson
reference2
Swing check valve
367
15-43> p. 31
Labor hours to install
1.1
15-43, p. 31
Labor costs at $2l.47/labor
hour
24
15-0, p. 2
Overhead costs at $13.3I/labor
hour
15
1-0, p. 5
Total direct costs
391
Administration: 10 percent of
total direct costs
39
1-0, p. 5
Taxes: 5 percent of equipment
cost
18
1-0, p. 6
Total indirect costs
72
Total installed cost
463
Annualized capital recovery cost3
75
CARD,
p. 3-18
Calculated as 0.16275 x (total installed cost), for an
rate of 10 percent and a 10-year recovery period.
interest
E-6
-------
TABLE E-5. DATA USED TO CALCULATE CONTROL DEVICE ANNUALIZED
COSTS1
(4th Quarter 1987 Dollars)
Item
Cost factor
Direct operating costs
Labor
$3,188 + ($11.65) x (16 person-hours) x (No. of scrubber
regenerations)3 °
Materials:
50 percent H2SO4
50 percent NaOH
Chlorine filters
Taxes
Freight
Compressed air
Disposal of ethylene glycol
($0.0702/lb) x (594 Ib/drum) x (No. of drums required) x (1.15)c-e
(Cost/lb) x (700 Ib/drum) x (No. of drums required) x (l.lSf • f' 8
($15/filter) x (No. of tank regenerations) x (No. of tanks)0 h
5 percent of materials cost
5 percent of materials cost
O1
-J
Indirect operating costs
Overhead
Property tax, insurance,
and administration
Capital recovery costs
(0.6) x (labor costs)
4 percent of total capital costs
(0. 16275) x (total capital costs)*
aNumber of scrubber regenerations = amount of EO to be treated divided by the conversion capacity of the
scrubber (See Example Calculation No. 6 in Section E.2)
*>The $3,188 is for visual inspection of the system 15 minutes per shift, 3 shifts per day, 365 days per year at
$11.65/person-hour. It was assumed that each regeneration of the scrubber solution would require two people
at 8 person-hours each, independent of scrubber size.
cNumber of scrubber tanks = scrubber conversion capacity divided by the conversion capacity of one tank
(2,000 pounds of EO). Number of tank regenerations = number of scrubber regenerations multiplied by the
number of scrubber tanks.
Each tank regeneration requires one 55-gallon drum of 50 percent t^SO^.
fifteen percent extra is allowed for spillage.
%ach tank regeneration requires 250 pounds of NaOH for neutralization.
SCost basis for 50 percent NaOH (350 pounds NaOH per drum):
fNo. of drums >9, cost/lb = $0.0110.
If No. of drums = 3 to 9, cost/lb = $0.0802.
If No. of drums <2, cost/lb = $0.0752.
hEach chlorine filter can dechlorinate approximately 200 gallons (one tank) of H2O; replace filter at each tank
.regeneration.
'The cost of 10 seconds of in-house compressed air per cycle is considered negligible.
JUnit cost of disposal was calculated by multiplying the total number of tank regenerations by the weight of a
tank at regeneration, approximately 4,845 Ib (see Example Calculation No. 3 in Section E.2.
If the total weight <42,000 Ib, disposal cost = (weight) x ($0.110/lb).
If the total weight 2:42,000 Ib, disposal cost = (weight) x ($0.068/lb).
^Assumes an interest rate of 10 percent and a 10-year recovery period.
E-7
-------
TABLE E-6. MISCELLANEOUS OPERATING COSTS1
Item description
Cost, 1987 $
Operating materials
1. 50 percent H2SO4, electrolyte-grade
2. 50 percent NaOH, industrial grade:
<2 drums
3-9 drums
>9 drums
3. Chlorine filters:
Filter housing
Filter
Installation
Shipping charges for waste disposal
Weight of solution for disposal:
<42,000 Ib (drums)
£42,000 Ib (bulk)
0.0702/lb
0.110/lb
0.0802/lb
0.0752/lb
41.50 each
15.00 each
20.00 each
i
0.110/lb
0.068/lb
E-8
-------
TABLE E-7. COST OF EtO ABATO^™
OXIDIZERS (F.O.B.)1'
(1987 Dollars)
CATALYTIC
Design flow rate,
m3/min (ft3/min)
28 (1,000)
84 (3,000)
168 (6,000)
252 (9,000)
336 (12,000)
Cost, $a b
48,000
81,000
112,000
148,000
189,000
aCosts in 1989 dollars were corrected to 1987 dollars using the
"Chemical Engineering" Plant Cost indices.
"Cost of replacement catalyst is $l,500/cell in 1989 dollars, or
approximately $l,240/cell in 1987 dollars.2'4'5
E-9
-------
E.I COSTS OF ACID-WATER SCRUBBERS
REFERENCES
1,
Telecon. Srebro, S., MRI, with D. Smith. Damas Corp.
June 20, 1986. Discussion about costs of the Damas |Tri-Phase
ethylene oxide scrubbers. I'
Telecon. Srebro, S., MRI, with D. Smith. Damas Corp.
December 12, 1989. Discussion about costs of Models 25 and
50 Damas acid/water scrubbers.
Beall, C., Meeting Minutes: Damas Corp. and Johnson &
Johnson. Midwest Research Institute. Raleigh, NC.
April 30, 1986. 9 p.
Telecon. Glanville, J., MRI, with C. Woltz. Union Carbide,
Inc. February 10, 1987. Discussion about flammability of EO
mixtures.
Telecon. Srebro, S., MRI, with M. Popescu. Johnson &
Johnson International. June 16, 1986. Discussion about the
Damas Tri-Phase ethylene oxide scrubbers.
E-10
-------
E.2 SAMPLE COST CALCULATIONS FOR ACID-WATER SCRUBBERS
-------
E.2 SAMPLE COST CALCULATIONS FOR ACID-WATER SCRUBBERS
Size, m3 (ft3)
Gutype
EO USE, kg Ob)
EO-EMTT, kg (Ib)
MEO-EMIT, Mg (tons)
Sterilization chambers at the facility ,
No. 1
19(667)
100
12,700 (28,000)
12,100 (26,600)
12.07 (13 .52)
No. 2
19 (667)
100
12,700 (28,000)
12,100 (26,600)
12.07 (13.52)
No. 3
34 (1,200)
12/88
540(1,200)
520 (1,140)
0.52 (0.58)
No. 4
37 (1,334)
100
21,000(46,000)
19,800 (43,700)
19.82 (22.20)
1 No. 5
37(1,334)
100
21,000(46,000)
19,800 (43,700)
19.82 (22.20)
Facility totals
EO-FAC, kg (Ib)
MEO-FAC, Mg (tons)
EO-TOT, kg (Ib)
MEO-TOT, Mg (ions)
CON-EM, Mg (iota)
REDUCE, Mg (tons)
64,310(141,740)
64.30 (72.02)
67,695 (149,200)
67.7 (74.6)
0.64 (0.70)
63.66 (70.15)
1. The size, gas type, and EO use are those for an actual
commercial sterilization facility represented in the EPA data
base. (See Section 3.1 of this report for a description of how
this data base was developed.) The other values were calculated
using the following assumptions:
a. EC-EMIT (Ib) = EO (Ib) emitted annually from an
individual sterilization chamber to the vacuum pump drain and to
the atmosphere. As shown in Section 3.4 of this documerit,
sterilizer vent emissions and vacuum pump drain emissions were
assumed to be 50 percent and 45 percent of EO use (Ib), !
respectively. Residual EO in the sterilized product prior to
aeration was assumed to be 5 percent of EO-USE (Ib). This
5 percent of the EO use is not included as part of EO-EMIT (Ib).
b. MEO-EMIT (Mg EO) = EO-EMIT (Ib)72,204.6
c. EO-FAC (Ib) and MEO-FAC (Mg) are the amount of EO
released annually by the facility to the vacuum pump drain and to
the atmosphere, i.e., the sum of EO-EMIT and the sum of MEO-EMIT,
respectively.
E-12
-------
d. EO-TOT (Ib) is the total amount of EO (Ib) used annually
by the facility, i.e., the sum of EO use. MEO-TOT (Mg) = EO-TOT
(lb)/2,204.6.
e. CON-EM (Mg) is the amount of EO that would be released
annually after control, i.e., MEO-TOT*(1-0.99)*0.95. Note that
the 5 percent residual EO in the sterilized product, which is
later released from the aeration room vent, is excluded from this
calculated emission estimate.
f. REDUCE (Mg) is the incremental amount of EO that would .
be reduced if controls are implemented, i.e., (MEO-FAC) -(CON-EM),.
2. For all calculations, a conversion efficiency of
99.0 percent was assumed for the scrubber.
3. Each tank of the scrubber initially holds 0.75 m3
(198 gal) H2O and 0.075 m3 (19.8 gal) H2SO4. The manufacturer
recommends that the tank be regenerated (i.e., drained, rinsed,
and refilled) after 907 kg (2,000 Ib) EO have been treated.
a. 0.075 m3 (19.8 gal) H2SO4 = 1.42 kg-mole H2SO4 (p =
1.834; MW = 98.08) 2NaOH + H2SO4 T Na2SO4 + 2H2O; 1.42 kg-mole
H2SO4 requires 2.84 kg mole NaOH to neutralize. Neutralization
will produce 2.84 kg-moles H2O and 1.42 kg mole Na2SO4. Use
50 percent (w/w) NaOH to neutralize; each 0.21-m3 (55-gal) drum
of 50 percent NaOH weighs 318 kg (700 Ib), i.e., 159 kg (350 Ib)
NaOH (MW = 40); need 2.84 kg-moles or 114 kg (250 Ib) NaOH to
neutralize.
b. C2H40 (EO) + H20 T C2H4(OH)2 (ethylene glycol); 907 kg
(2,000 Ib) EO = 20.51 kg-moles EO (MW = 44.1).
c. At 99 percent conversion, yield is 20.365 kg-moles or
1.14 m3 (301 gal) ethylene glycol (EG) (MW = 62.07; p = 1.1088).
d. At 99 percent conversion, 20.365 kg-moles H2O have
reacted. 41.64 kg moles H2O originally available (MW = 18;
p = 1); 21.275 kg-moles or 0.38 m3 (100 gal) H2O remain
unreacted.
e. Weight of neutralized solution per tank: 1.42-kg mole
Na2S04 = 202 kg Na2SO4 (MW = 142.04); 2.84 kg-moles H2O (from
neutralization) = 51 kg (112 Ib) H2O; 250 Ib (113 kg) H2O = from
50 percent NaOH solution; 0.38 m3 (100 gal) unreacted H2O
E-13
-------
- 378 kg H20 (833 lb); 1.14 m3 (301 gal)EG = 1,264 kg
(2,786 lb) EG; total wt = 2,008 kg = 4,427 lb.
f. Solution is 63 percent (w/w) EG. Add about 0.19 m3
(50 gal) rinse water for each tank = 189 kg (416 lb); total wt
(+ rinse H2O) = 2,198 kg (4,844 lb); total gal (+ rinse1 H2O)
- 1.87 m3 (495 gal) = nine 0.21 m3 (55-gal) drums; wt per 0.21 m3
(55-gal) drum = 244 kg (538 lb). '
4. Find scrubber model and cost from Table E-l, based on
the sum of the volumes of the two largest chambers at the
facility:
Chambers 4 and 5 75 m3 (2,668 ft3) Model 600 $157,500
5. Because at least one chamber uses 100 percent EO,
explosion-proof valves are necessary.
6. Find number of regenerations of scrubber required per
year: ;
a. Number of scrubber tanks = scrubber model/100 = 6
(scrubber consists of modular tanks). !
b. Conversion capacity of scrubber = (No. of tanks) x
2,000 lb = 12,000 lb.
c. Number of scrubber regenerations = EO-FAC (lb) j/12,000,
i.e., the amount of EO (lb) to be treated per year divided by the
conversion capacity of the scrubber.
141,700/12,000 = 11.81 scrubber regenerations/yr
d. Number of tank regenerations = (No. of scrubber
regenerations) x (No. of tanks per scrubber) \
- (11.81) X (6) = 70.87.
7. Cost of chlorine filter housing = (41.50) x (No. of
tanks) = $(41.50) x (6) = $249.
8. Installation costs:
a. Scrubber installation = (0.5) x (cost of
scrubber) *= $78,750
b. Chlorine filter housing installation = (20) x i(No. of
tanks) = $120. !
9. The incremental capital costs of manifolding are
presented in Table E-14 of this report. '•
E-14
-------
10. Vacuum pumps. A closed-loop recirculating water vacuum
pump is required on each of the five chambers. The cost of
modifying the first vacuum pump is included in the cost of the
scrubber; the cost of modifying the other four vacuum pumps is
$4,935 each. .
11. Calculate direct operating costs:
a. Labor = 3,177 + (11.60)x(16) x (No. of regenerations).
The $3,177 is for general inspection of the system
15 minutes/shift, 3 shifts/d, 365 d/yr at $11.60/person-hour.
For the purposes of these cost analyses, it was assumed that each
regeneration of the scrubber would require 2 people at 8 person-
hours each, independent of scrubber size. System inspection was
also assumed to be independent of scrubber size.
b. Sulfuric acid (50 percent H2SO4-electrolyte grade).
Assumed: 1 55-gal drum of 50 percent H2S04, i.e., 19.4 gal
H2SO4' Per scrubber tank.
No. of drums required = No. of tank regenerations = (No. of
scrubber regenerations) x (No. of tanks per scrubber) = 70.87
Cost of acid = (No. of drums) x (594 Ib/drum) x ($0.069/lb)
c. Caustic (50 percent NaOH-industrial grade). First, the
unit cost of NaOH was calculated.
NaOH required per year = [No. of tank regenerations] x [NaOH
(Ib) required per tank] = 70.87 x 250 = 17,718 Ib/yr
Total drums/yr required by facility = total NaOH (lb)/350 Ib
per drum; total drums = 50.6
If total drums >9, cost/lb = 0.0738
If total drums = 3 to 9, cost/lb = 0.0787
If total drums = <2, cost/lb - 0.108
Cost of caustic = (No. of drums) x (cost/lb) x (700 Ib/drum)
d. Cost of chlorine filters. Each filter can dechlorinate
f200 gal H2O (or 1 scrubber tank); replace at each scrubber
regeneration.
Cost = (No. of scrubber regenerations) x (No. of tanks) x
($15/filter)
E-15
-------
I
e. Disposal. Unit cost of disposal was calculated by
multiplying the total number of tank regenerations by the weight
of a tank at the time of regeneration, including rinse water
(see 3.f). ;
Total Wt = 70.87 x 4,844 Ib/tank =343,943 Ib/yr !
If total wt <42,000 Ib, disposal cost = wt (Ib) x
($0.096/lb)
If total wt >42,000 Ib, disposal cost = wt (Ib) x
($0.059/lb) I
f. Compressed air. The cost of 10 seconds of in-house air
per cycle was considered negligible and was not computed for
these cost analyses.
12. The capital and annualized costs are reportediin
Tables E-3 and E-5.
REFERENCES
1.
2.
3.
Memorandum. Srebro, S., to D. Markwordt. EPA/CPB.1 Cost
Effectiveness of Reducing Ethylene Oxide Emissions ifrom
Sterilizer Vents and Associated Vacuum Pump Drains.'
March 21, 1991.
Richardson Engineering Services, Inc. Process Plant
Construction Estimating Standards. 1984.
Neveril, R., Capital and Operating Costs of Selected Air
Pollution Control Systems. CARD, Inc. Niles, IL.
Publication No. EPA-450/5-80-002. December 1978.
E-16
-------
E.3 AERATION ROOM COST ANALYSIS
-------
TABLE E-8.
AERATION ROOM GAS/SOLID REACTANT
COST ANALYSIS
CONTROL
1 2.00
2.98
II 4<8°
6.00
6.00
1 6'°°
6.00
7.50
II 9.00
1 12-°°
| 16.80
17.80
1 27.36
1 32.40
|j 32.40
32.40
61.34
64.00
64.80
65.00
70.00
93.60
100.80
105.00
110.00
111.00
113.00
115.00
126.00
160.80
162.00
185.25
188.00
192.00
194.00
195.00
200.00
201.60
264.00
291.60
1 312.00
312.00
1 324.00
1
1
1
1
1
1
1
1
I
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
I
1
1
1
1
1
1
1
1
1
1
1
1
1
1
I
i COST CELLS
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
NUM 1000
1
1
1
1
1
• 1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
NUM 3000
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
CAT FOB
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
TCC
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
TAC
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
ARRED
0.0000202
0.0000269
0.0000401
0.0000647
0.0000808
0.0000808
0.0000808
0.0000808
0.0001010
0.0001212
0.0001617
0.0002263
0.0002398
0.0003686
0.0004365
0.0004365
0.0004365
0.0008264
0.0008622
1
0.0008730
0.0008757
0.0009430
0.0012610
0.0013580
0.0014145
0.0014819
0.0014954
0.0015223
0.0015493
0.0016975
0.0021663
0.0021824
0.0024957
0.0025327
0.0025866
0.0026135
0.0026270
0.0026944
0.0027159
0.0035566
0.0039284
0.0042032
0.0042032
0.0043649
1500000000
1100000000
750000000
460000000
370000000
370000000
370000000
370000000
300000000
250000000
190000000
130000000
130000000
81000000
69000000
69000000
69000000
36000000
35000000
34000000
34000000
32000000
24000000
22000000
21000000
20000000
20000000
20000000
19000000
18000000
14000000
14000000
12000000
12000000
12000000
11000000
11000000
11000000
11000000
8400000
7600000
7100000
7100000
6900000
E-18
-------
TABLE E-8. (continued)
EO TOT
388.80
390.00
395.00
432.00
456.00
504.00
520.00
557.30
562.00
581.00
686.00
804.00
850.00
875.00
1092.00
1155.00
1231.00
1300.00
1334.00
1521.00
1600.00
1714.00
1714.00
1750.00
1750.00
1800.00
1944.00
1965.00
1980.00
2016.00
2142.00
2254.00
2280.00
2376.00
2419.20
2450.00
2527.00
2599.20
3120.00
3135.00
3240.00
3640.00
3755.00
4200.00
HUM CELLS
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
COST CELLS
0
0
16000
0
0
0
0
0
0
0
0
0
0
0
0
16000
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
MUM 1000
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
NUM 3000
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0'
0
0
0
0
0
0
0
0
CAT.FOB
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
TCC
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
63600
TAG
30000
30000
33700
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
33700
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
30000
ARRED
0.0052378
0.0052540
0.0053214
0.0058198
0.0061432
0.0067898
0.0070054
0.0075079
0.0075712
0.0078271
0.0092417
0.0108314
0.0114511
0.0117879
0.0147112
0.0155600
0.0165838
0.0175134
0.0179714
0.0204907
0.0215549
0.0230907
0.0230907
0.0235757
0.0235757
0.0242493
0.0261892
0.0264721
0.0266742
0.0271592
0.0288567
0.0303655
0.0307158
0.0320091
0.0325911
0.0330060
0.0340433
0.0350160
0.0420321
0.0422342
0.0436487
0.0490375
0.0505867
0.0565817
CEFF
5700000
5700000
6300000
5200000
4900000
4400000
4300000
4000000
4000000
3800000
3200000
2800000
2600000
2500000
2000000
2200000
1800000
1700000
1700000
1500000
1400000
1300000
1300000
1300000
1300000
1200000
1100000
1100000
1100000
1100000
1000000
1000000
1000000
940000
920000
910000
880000
860000
710000
710000
690000
610000
590000
530000
E-19
-------
TABLE E-8. (continued)
I
EO_TOT |NUM_CELLS|COST CELLS! NUM iooo| NUM aoool CAT FO
' i - i - i -
4200.00
4286.90
4320.00
4368.00
4667.00
4860.00
5016.00
5088.00
5189.00
5250.00
5258.00
5739.00
5850.00
6000.00
6048.00
6176.00
6451.00
6840.00
6900.00
7194.00
7350.00
7387.00
8390.00
8400.00
8736.00
9676.00
10002.00
10613.00
10800.00
11016.00
11400.00
11440.00
11547.00
11984.00
12020.00
12249.00
13000.00
13000.00
13059.00
14352.00
14400.00
14860.00
14862.00
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
I
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
3
0
0
0
0
0
0
16000
0
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
0
16000
16000
16000
32000
32000
32000
32000
32000
32000
32000
32000
32000
32000
32000
48000
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
3 48000 1 o 44200
TCC | TAG
63600
63600
63600
63600
63600
63600
82000
63600
82000
82000
82000
82000
82000
82000
82000
82000
82000
82000
82000
82000
82000
82000
82000
82000
82000
82000
63600
82000
82000
82000
21000
21000
21000
21000
21000
21000
21000
21000
21000
21000
21000
43000
43000
30000
30000
30000
30000
30000
30000
33700
30000
33700
33700
33700
33700
33700
33700
33700
33700
33700
33700
33700
33700
33700
33700
33700
33700
33700
33700
33700
33700
33700
33700
41700
41700
41700
41700
41700
41700
41700
41700
41700
41700
41700
46100
46100
^===
ARRED
0.0565817
0.0577524
0.0581983
0.0588450
0.0628730
0.0654731
0.0675747
0.0685447
0.0699053
0.0707271
0.0708349
0.0773148
0.0788102
0.0808310
0.0814776
0.0832020
0.0869068
0.0921473
0.0929556
0.0969164
0.0990180
0.0995164
0.1130287
0.1131634
1
0.1176899
0.1303534
0.1347453
0.1429765
0.1454958
0.1484057
0.1535789
0.1541178
0.1555592
0.1614464
0.1619314
0.1650165
0.1751338
0.1751338
0.1759286
0.1933477
0.1939944
0.2001914
0.2002184
CEFF 1
530000
520000
520000
510000
480000 |
460000
500000
440000
480000
480000
480000
440000
430000
420000
410000
410000
390000
370000
360000
350000
340000
340000
300000
300000
290000
260000
250000
240000
230000
230000
270000
270000
270000
260000
260000
250000
240000
240000
240000
220000
210000
230000
230000
E-20
-------
TABLE E-8. (continued)
EO TOT
15040.00
15600.00
15724.00
16426.00
16956.00
18000.00
18057.00
18268.00
18963.00
19090.00
21135.00
22560.00
23954.00
24000.00
24082.00
26549.00
29500.00
29700.00
30034.00
30570.00
32040.00
34034.00
35040.00
36231.00
38799.00
39105.00
40188.00
41320.00
43050.00
45804.00
50000.00
52547.36
52950.00
54391.00
55567.00
56860.00
64044.00
66503.00
78580.00
88000.00
93450.00
100524.00
108408.00
109897.00
HUM CELLS
2
1
2
1
2
2
4
1
2
2
2
2
2
2
2
4
4
4
4
5
4
5
5
5
5
4
5
9
8
3
7
7
7
7
8
9
6
9
1
11
11
13
18
6
COST CELLS
32000
16000
32000
0
32000
32000
64000
32000
48000
48000
48000
48000
48000
48000
48000
64000
64000
64000
64000
80000
64000
80000
80000
80000
80000
64000
80000
144000
128000
48000
112000
112000
112000
112000
128000
144000
96000
144000
0
176000
176000
208000
288000
96000
NUMJOOO
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
0
0
0
0
1
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
1
0
0
NUM_3000
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
1
1
1
0
1
1
1
0
1
1
1
1
1
I
1
1
0
1
1
1
2
1
CATJFOB
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
44200
74800
44200
74800
74800
74800
74800
44200
74800
74800
74800
44200
74800
74800
74800
74800
74800
74800
74800
74800
44200
74800
74800
11900
150000
74800
TCC
121000
82000
121000
63600
121000
121000
165000
121000
143000
143000
143000
143000
143000
143000
143000
165000
165000
165000
165000
231000
165000
231000
231000
231000
231000
165000
231000
319000
297000
143000
275000
275000
275000
275000
297000
319000
253000
319000
63600
363000
363000
470000
626000
253000
TAG
41700
33700
41700
30000
41700
41700
50500
41700
46100
46100
46100
46100
46100
46100
46100
50500
50500
50500
50500
75300
50500
75300
75300
75300
75300
50500
75300
93300
88800
46100
84300
84300
84300
84300
88800
93300
79800
93300
30000
102000
102000
141000
184000
79800
ARRED
0.2026163
0.2101606
0.2118311
0.2212883
0.2284284
0.2424930
0.2432609
0.2461034
0.2554663
0.2571773
0.2847272
0.3039245
0.3227043
0.3233240
0.3244286
0.3576637
0.3974190
0.4001134
0.4046130
0.4118339
0.4316375
0.4585003
0.4720530
0.4880979
0.5226936
0.5268160
0.5414060
0.5566561
0.5799623
0.6170638
0.6735916
0.7079092
0.7133335
0.7327464
0.7485893
0.7660083
0.8627900
0.8959172
1.0586165
1.1855212
1.2589427
1.3542424
1.4604543
1.4805139
CEFF
210000
160000
200000
140000
180000
170000
210000
170000
180000
180000
160000
150000
140000
140000
140000
140000
130000
130000
120000
180000
120000
160000
160000
150000
140000
96000
140000
170000
150000
75000
130000
120000
120000
120000
120000
120000
92000
100000
28000
86000
81000
100000
130000
54000
E-21
-------
TABLE E-8. (continued)
EOJTOT
125084.00
134956.00
149000.00
162287.00
184511.60
184766.00
197260.00
215000.00
240030.00
283998.00
NUM_CELLS
9
19
10
10
11
11
22
15
16
17
COST_CELLS
144000
304000
160000
160000
176000
176000
352000
240000
256000
272000
NUMJOOO
0
0
0
0
0
0
0
1
1
0
NUM 3000
1
2
1
1
1
1
2
1
1
2
CAT FOB
74800
150000
74800
74800
74800
74800
150000
119000
119000
150000
TCC
319000
648000
340000
340000
363000
363000
713000
514000
537000
603000
TAG
93300
189000
97500
97500
102000
102000
201000
150000
155000
179000
ARRED
1.6851106
1.8181045
2.0073029
2.1863031
2.4857092
2.4891364
2.6574535
2.8964438
3.2336437
3.8259732
CEFF
55000
100000
49000
45000
41000
41000
76000
52000
48000
47000
E-22
-------
Notes for Table E-8
1. The field "AROOM" indicates facilities used in the model with
an "*".
2. The field "EOJTOT" is the annual EP use at the facility.
3. The field "NUM CELLS" is the number of aeration cells
assigned to that facility.
4. The field "COST CELLS" gives the cost of aeration cells.
5. The field "NUM_1000" is the number of 1,000 ft3/min control
units.
6. The field "NUM_3000" is the number of 3,000 ft3/min control
units.
7. The field "SAFE FOB" is the capital cost (FOB) of a gas/solid
reactor control.
8. The field "TCC" gives the total capital cost for the
facility.
9. The field "TAG" is the total annual control cost for the
facility.
10. The field "ARRED" is the annual emission reduction (mg).
11. The field "CEFF" gives the cost effectiveness ($/Mg).
E-23
-------
TABLE E-9.
AERATION ROOM CATALYTIC OXIDATION CONTROL COST
ANALYSIS ;
EO_TOT
1.50
2.00
2.98
4.80
6.00
6.00
6.00
6.00
7.50
9.00
12.00
16.80
17.80
27.36
32.40
32.40
32.40
61.34
64.00
64.80
65.00
70.00
93.60
100.80
105.00
110.00
111.00
113.00
115.00
126.00
160.80
162.00
185.25
188.00
192.00
194.00
195.00
200.00
201.60
264.00
291.60
312.00
312.00
324.00
NUM.CELLS
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
I
COST_CELLS
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
NUMJOOO
1
1
1
1
1 ,
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
NUMJOOO
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
SAFE_FOB
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
TCC
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
.27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
TAG
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
ARRED
0.0000202
0.0000269
0.0000401
0.0000647
0.0000808
0.0000808
0.0000808
0.0000808
0.0001010
.0.0001212
0.0001617
0.0002263
0.0002398
0.0003686
0.0004365
0.0004365
0.0004365
0.0008264
0.0008622
0.0008730
0.0008757
0.0009430
0.0012610
0.0013580
0'.0014145
0.0014819
0.0014954
0.0015223
0.0015493
0.0016975
0.0021663
I
0.0021824
0.0024957
0.0025327
0.0025866
0.0026135
0.0026270
0.0026944
0.0027159
0.0035566
0.0039284
0.0042032
0.0042032
0.0043649
CEFF
460000000
350000000
230000000
140000000
120000000
120000000
120000000
120000000
92000000
77000000
58000000
41000000
39000000
25000000
21000000
21000000
21000000
11000000
11000000
11000000
11000000
9900000
7400000
6800000
6600000
6300000
6200000
6100000
6000000
5500000
4300000
4300000
3700000
3700000
3600000
3600000
3500000
3500000
3400000
2600000
2400000
2200000
2200000
2100000
E-24
-------
TABLE E-9. (continued)
EG TOT
388.80
390.00
395.00
432.00
456.00
504.00
520.00
557.30
562.00
581.00
686.00
804.00
850.00
875.00
1092.00
1155.00
1231.00
1300.00
1334.00
1521.00
1600.00
1714.00
1714.00
1750.00
1750.00
1800.00
1944.00
1965.00
1980.00
2016.00
2142.00
2254.00
2280.00
2376.00
2419.20
2450.00
2527.00
2599.20
3120.00
3135.00
3240.00
3640.00
3755.00
4200.00
NUMjCELLS
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
I
1
1
1
1
1
1
1
1
1
1
1
1
COST CELLS
0 .
0
16000
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
NUMJOOO
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
I
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
NUM_3000
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
- 0
0
0
0
0
0
0
0
0
SAFE_FOB
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
' 17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
TCC
27500
27500
45900
27500
;27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
27500
TAG
9300
9300
13000
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
9300
ARRED
0.0052378
0.0052540
0.0053214
0.0058198
0.0061432
0.0067898
0.0070054
0.0075079
0.0075712
0.0078271
0.0092417
0.0108314
0.0114511
0.0117879
0.0147112
0.0155600
0.0165838
0.0175134
0.0179714
0.0204907
0.0215549
0.0230907
0.0230907
0.0235757
0.0235757
0.0242493
0.0261892
0.0264721
0.0266742
0.0271592
0.0288567
0.0303655
0.0307158
0.0320091
0.0325911
0.0330060
0.0340433
0.0350160
0.0420321
0.0422342
0.0436487
0.0490375
0.0505867
0.0565817
CEFF
1800000
1800000
2400000
1600000
1500000
1400000
1300000
1200000
1200000
1200000
1000000
860000
810000
790000
630000
600000
560000
530000
520000
450000
430000
400000
400000
390000
390000
380000
360000
350000
350000
340000
320000
310000
300000
290000
290000
280000
270000
270000
220000
220000
210000
190000
180000
160000
E-25
-------
TABLE E-9. (continued)
| EOTOT
4200.00
4286.90
4320.00
4368.00
4667.00
4860.00
5016.00
5088.00
5189.00
5250.00
5258.00
5739.00
5850.00
6000.00
I
6048.00
6176.00
1
6451.00
6840.00
1
6900.00
7194.00
7350.00
7387.00
8390.00
8400.00
8736.00
9676.00
10002.00
10613.00
10800.00
11016.00
11400.00
11440.00
11547.00
11984.00
12020.00
12249.00
13000.00
13000.00
13059.00
14352.00
14400.00
14860.00
14862.00
15040.00
NUM.CELLS
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
I
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2 _
3
3
2
COST CELLS
0
0
0
0
0
0
16000
0
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
16000
0
16000
16000
16000
32000
32000
32000
32000
32000
32000
32000
32000
32000
32000
32000
48000
48000
32000
MUM 1000
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
r
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
MUM 3000
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
SAFE FOB
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
TCC
27500
27500
27500
27500
27500
27500
45900
27600
45900
45900
45900
45900
45900
45900
45900
45900
45900
45900
45900
45900
45900
45900
45900
45900
45900
45900
27500
45900
45900
45900
74600
74600
74600
74600
74600
74600
74600
74600
74600
74600
74600
94800
94800
74600
TAC
9300
9300
9300
9300
9300
9300
13000
9300
13000
13000
13000
13000
13000
13000
13000
13000
13000
13000
13000
13000
13000
13000
13000
13000
13000
13000
9300
13000
13000
13000
18800
18800
18800
18800
18800
18800
18800
18800
18800
18800
18800
22900
22900
18800
ARRED
0.0565817
0.0577524
0.0581983
0.0588450
0.0628730
0.0654731
0.0675747
0.0685447
0.0699053
0.0707271
0.0708349
0.0773148
0.0788102
0.0808310
0.0814776
0.0832020
0.0869068
0.0921473
0.0929556
0.0969164
0.0990180
0.0995164
0.1130287
0.1131634
0.1176899
0.1303534
0.134:7453
0.1429765
0.1454958
0.1484057
0.1535789
0.1541178
0.1555592
0.1614464
0.1619314
0.1650165
0.1751338
0.1751338
0.1759286
0.1933477
0.1939944
0.2001914
0.2002184
0.2026163
CEFF
160000
160000
160000
160000
150000
140000
190000
140000
190000
180000
180000
170000
160000
160000
160000
160000
150000
140000
140000
130000
130000
130000
120000
110000
110000
100000
69000
91000
89000
88000
120000
120000
120000
120000
120000
110000
110000
110000
110000
97000
97000
110000
110000
93000
E-26
-------
TABLE E-9. (continued)
EO_TOT
15600.00
15724.00
16426.00
16956.00
18000.00
18057.00
18268.00
18963.00
19090.00
21135.00
22560.00
23954.00
24000.00
24082.00
26549.00
29500.00
29700.00
30034.00
30570.00
32040.00
34034.00
35040.00
36231.00
38799.00
39105.00
40188.00
41320.00
43050.00
45804.00
50000.00
52547.36
52950.00
54391.00
55567.00
56860.00
64044.00
66503.00
78580.00
88000.00
93450.00
100524.00
108408.00
109897.00
125084.00
NUM.CELLS
1
2
1
2
2
4
1
2
2
2
2
2
2
2
4
4
4
4
5
4
5
5
5
5
4
5
9
8
3
7
7
7
7
8
9
6
9
1
11
11
13
18
6
9
COST.CELLS
16000
32000
0
32000
32000
64000
32000
48000
48000
48000
48000
48000
48000
48000
64000
64000
64000
64000
80000
64000
80000
80000
80000
80000
64000
80000
144000
128000
48000
112000
112000
112000
112000
128000
144000
96000
144000
0
176000
176000
208000
288000
96000
144000
NUMJOOO
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
0
0
0
0
1
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
1
0
0
0
NUM_3000
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
1
1
1
0
1
1
1
0
1
1
1
1
1
1
1
1
0
1
1
1
2
1
1
SAFE_FOB
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
17100
37000
17100
37000
37000
37000
37000
17100
37000
37000
37000
17100
37000
37000
37000
37000
37000
37000
37000
37000
17100
37000
37000
54100
73900
37000
37000
TCC
45900
74600
27500
74600
74600
115000
74600
94800
94800
94800
94800
94800
94800
94800
115000
115000
115000
115000
167000
115000
167000
167000
167000
167000
115000
167000
248000
228000
94800
208000
208000
208000
208000
208000
248000
188000
248000
27500
289000
289000
357000
489000
188000
248000
TAC
13000
18800
9300
18800
18800
27000
18800
22900
22900
22900
22900
22900
22900
22900
27000
27300
27300
27300
46400
27400
46400
46400
46400
46400
28000
46400
62900
58900
24300
54800
54800
54800
54800
58900
32900
50800
62900
13200
71900
72300
88600
124000
52900
66300
ARRED
0.2101606
0.2118311
0.2212883
0.2284284
0.2424930
0.2432609
0.2461034
0.2554663
0.2571773
0.2847272
0.3039245
0.3227043
0.3233240
0.3244286
0.3576637
0.3974190
0.4001134
0.4046130
0.4118339
0.4316375
0.4585003
0.4720530
0.4880979
0.5226936
0,5268160
0.5414060
0.5566561
0.5799623
0.6170638
0.6735916
0.7079092
0.7133335
0.7327464
0.7485893
0.7660083
0.8627900
0.8959172
1.0586165
1.1855212
1.2589427
1.3542424
1.4604543
1.4805139
1.6851106
CEFF
62000
89000
42000
82000
78000
110000
76000
90000
89000
80000
75000
71000
71000
71000
75000
69000
68000
67000
110000
63000
100000
98000
95000
89000
53000
86000
110000
100000
39000
81000
77000
77000
75000
79000
43000
59000
70000
12000
eiooo
57000
65000
85000
36000
39000
E-27
-------
TABLE E-9. (continued)
EO TOT
134956.00
149000.00
162287.00
184511.60
184766.00
197260.00
215000.00
240030.00
283998.00
NUM CELLS
19
10
10
11
11
22
15
16
17
COST CELLS
304000
160000
160000
176000
176000
352000
240000
256000
272000
NUM 1000
0
0
0
0
0
0
1
1
0
NUM 3000
2
1
1
1
1
2
1
1
2
SAFE_FOB
73900
37000
37000
37000
37000
73900
54100
54100
73900
TCC
510000
268000
268000
289000
289000
570000
397000
418000
469000
TAG
129000
72100
73100
79100
79100
144000
105000
111000
130000
ARRED
1.8181045
2.0073029
2.1863031
2.4857092
2.4891364
2.6574535
2.8964438
3.2336437
3.8259732
CEFF
71000
36000
33000
32000
32000
54000
36000
34000
34000
E-28
-------
E.3 AERATION ROOM COST ANALYSIS
Notes for Table E-9
1. The field "AROOM" indicates facilities used in the model
with an "*".
2. The field "EO_TOT" is the annual EO use at the facility.
3. The field "NUM_CELLS" is the number of aeration cells
assigned to that facility.
4. The field "COST_CELLS" gives the cost of aeration cells.
5. The field "NUMJLOOO" is the number of 1,000 ft3/min control
units.
6. The field MNUM_3000" is the number of 3,000 ft3/min control
units.
7. The field "CATJFOB" is the capital cost (FOB) of a catalytic
oxidation control.
8. The field "TCC" gives the total capital cost for the
facility.
9. The field "TAG" is the total annual control cost for the
facility.
10. The field "ARRED" is the annual emission reduction (mg).
11. The field "CEFF" gives the cost effectiveness ($/Mg).
REFERENCES
1. Letter and attachments from Olson, C., Donaldson Company,
Inc., to S. Srebro. MRI. March 23, 1989. Capital and
operating costs of 1,000 ft3/min EtO Abater™ catalytic
oxidizer.
2. Telecon. Srebro, S., MRI, with C. Olson. Donaldson
Company, Inc. April 4, 1989. Discussion about costs of EtO
Abater™.
3. Telecon. Nicholson, R., MRI, with C. Olson. Donaldson,
Company, Inc. May 12 and June 13, 1988. Costs of EtO
Abaters™.
4. Chemical Engineering. Economic Indicators. April 25, 1988.
p. 9.
E-29
-------
5. Chemical Engineering. Economic Indicators. June 1989,
p. 224. I
E-30
-------
E.4 EXAMPLE FACILITY CALCULATIONS
-------
TABLE E-10.
CAPITAL AND ANNUAL COSTS OF INSTALLING SCRUBBERSJ
(4th Quarter 1987 Dollars)
[tern
Cost
CAPITAL COSTSa
1. Installed equipment costs
a. Acid/water scrubber'
b. Explosion-proof valves for scrubber0
c. Chlorine filter house"
d. Purchased equipment costs, total
e. Installation of scrubber6
f. Installation of chlorine filters
g. Taxes: 5 percent of equipment cost
h. Freight: 5 percent of equipment cost
i. Vacuum pump
j. Manifolding of chambers (includes check valve)8
k. Subtotal Capital Costs
1. Contingencies'1
TOTAL CAPITAL INVESTMENT
98,900
N/AC
166
99,100
49,500
80
4,960 1
4,960
0
463
159,000
15,900
175,000
B. ANNUAL COSTSa
1. Direct operating costs
a. Labor1
b. Materials
50 percent I^SO^
50 percent NaOHk
Chlorine filters1
Taxes: 5 percent of materials cost
Freight: 5 percent of materials cost
c. Water"
d. Electricity11
e. Compressed air°
f. Disposal of ethylene glycolP
2. Indirect operating costs
a. Overhead: 0.60 x
b. Property tax, insurance, and administration^
c. Capital recovery costsr
TOTAL ANNUAL COSTS
3,540
364
350
117
42
42
0
124
0
4,040
2,120
7,000
28,500
46,200 i
C. COST EFFECTIVENESS
1. Emission reduction, Mg EO/yr (tons EO/yr)
2. Cost effectiveness, 1987, $/Mg ($/ton EO)
6.8 (7.5)
6,800 (6,200)
E-32
-------
TABLE E-10. (continued)
aCosts rounded to three significant figures.
"Model 400 based on largest chamber size of 1,000 ft3.
cNot applicable.
dOne per tank at $41.50 each (four tanks-Model 400).
^ifty percent of scrubber cost.
One at $5,170. The cost of the first vacuum pump is included in the installation cost of the scrubber.
jjSee Table E-4. One chamber costed for a check valve at $463.
. Assumed to be 10 percent of subtotal capital costs to account for uncertainties in the capital cost estimates.
'Labor was calculated for 0.25 person-hours/shift, 3 shifts/day, 365 days/year for system inspection and
. 16 person-hours for each regeneration of the scrubber at $11.65/person-hour.
JThe cost of acid is calculated, (UNC_FACH(2,000)x(594)x($0.070)xl.l5. (15 percent extra is for
spillage.) (UNCJFAC) is equal to uncontrolled emissions (Ib) from vent and drain at baseline.)
kThe cost of caustic is calculated, No. drum = (UNC_FAC)-K2,000)x (250)-s-(350). No. drum = 5.4;
therefore, unit cost = $0.0802. Total cost = (No.drums)x(700)x(0.0802)xl.l5(15 percent extra for
spillage).
'Chlorine filter cost is (UNC_FAC)x(15)-=-(2,000).
"Calculated as (scrubber model)x(2)x(UNC_FAC)/[(2,000)x(No. of tanks)]x(0.25/l,000).
nSee proceeding notes for calculation methodology.
°The cost of 10 seconds of house-supplied compressed air per cycle was considered negligible.
PDisposal cost is (UNC_FAC)H-(2,000)x(4,845)x(0.110).
^Calculated as 4 percent of total capital costs.
rCalculated as (0.16275)x(total capital costs) for an interest rate of 10 percent and a 10-year recovery period.
E-33
-------
TABLE E-ll.
CAPITAL AND ANNUAL COSTS OF INSTALLING SCRUBBERS
TO CONTROL CHAMBER EXHAUST VENTS2 ;
(4th Quarter 1987 Dollars)
Item
Cost
A. CAPITAL COSTS8
1. Installed equipment costs
a. Acid/water scrubber"
b. Chlorine filter house0
c. Purchased equipment costs, total"
d. Installation of chlorine filters"
e. Sales tax: 5 percent of equipment cost"
f. Freight: 5 percent of equipment cost"
g.' Manifolding of ventse
TOTAL CAPITAL INVESTMENT^
44,900
42
44,900
20
2,240
2,240
8,410
117,000
B. ANNUAL COSTS3
1. Direct operating costs
a. LaborS
b. Materials
(1) 50 percent H2S04h
(2) 50 percent NaOH*
(3) Chlorine filters)
(4) Taxes: 5 percent of materials cost
(5) Freight: 5 percent of materials cost
c. Water*
d. Electricity1
e. Compressed air™
f. Disposal of ethylene glycoP
2. Indirect operating costs
a. Overhead: 0.60 x labor
b. Property tax, insurance, and administration0
c. Capital recovery costs?
TOTAL ANNUAL COSTS
3,310
31
42
10
4
4
0
82
0
350
1,990
4,680
19,000
29,500
C. COST EFFECTIVENESS8
1. Emission reduction, Mg EO/yr (ton EO/yr)
2. Cost effectiveness, 1987, $/Mg EO ($/ton EO)
0.71 (0.78)
30,000 (27,000)
E-34
-------
TABLE E-ll. (continued)
"Capital and annual costs rounded to three significant figures. Cost effectiveness rounded to two significant
figures.
''Facility has four chambers. Therefore, costed for a 6,000 ft^/min scrubber.
cOne per tank at $41.50 each.
^Capital equipment cost.
eSee Section E.5 for a-detailed summary of manifolding costs.
'Capital costs were increased by a factor of 2.2 (except manifolding costs) to account for additional capital
expenditures necessary to install the control. The total capital costs were calculated as (2.2 x {sum of items
with superscript d]) + manifolding costs.
SLabor was calculated for 0.25 person-hours/shift, 3 shifts/day, 365 days/year for system inspection and
16 person-hours for each regeneration of the scrubber at $11.65/person-hour.
hThe cost of acid is calculated, (UNC_RCE) -i- (2,000) x (594) x ($0.070) x 1.15. (Extra 15 percent for
.spillage.) (UNC_RCE is equal to uncontrolled emissions (Ib) from rear chamber exhaust vents.)
'The cost of caustic is calculated, No. drum = (UNC-RCE) -5- (2,000) x (250) -^ (350). No. drum = 0.6;
therefore, unit cost = $0.110. Total cost = (No. drums) x (700) x ($0.11) x 1.15 (15 percent extra for
.spillage).
JChlorine filter cost is (UNC_RCE/2,000) x 15.
jk^lculated as (scrubber model) x (2) x (UNCJRCE/2,000) x (0.25/1,000).
'See proceeding notes for calculation methodology.
"The cost of 10 seconds of house-supplied compressed air per cycle was considered negligible.
DDisposal cost is (UNC_RCE) -=- (2,000) x (4,845) x (0.110).
°Calculated as 4 percent of total capital investment.
PCalculated as (0.16275) x (total capital investment) (i.e., an interest rate of 10 percent and a 10-year recovery
period). '
E-35
-------
TABLE E-12. CAPITAL AND ANNUAL COSTS OF GAS/SOLID REACTOR
TO CONTROL AERATION UNITS AT AN EXAMPLE FACILITY3
(4TH QUARTER 1987 Dollars)
Item
A. CAPITAL COSTS
1. Installed Equipment Costs
x. Gas/solid reactor
b. Installation of gas/solid reactor*
c. Taxes: 5 percent of equipment cost
d. Freight: 5 percent of equipment cost
e. Aeration units, installed cost"
f. Manifolding of aeration units0
g. Subtotal capital costs
h. Contingencies
TOTAL CAPITAL COSTS
B. ANNUAL COSTS
1. Direct operating Costs
a. Maintenance labor6
b. Maintenance materials'
c. Reactant replacement^
d. Labor for reactant replacement"
e. Electricity1
f. Disposal of reactant!
2. Indirect Operating Costs
a. Overhead^
b. Property tax, insurance, and administration'
c. Capital recovery costs™
TOTAL ANNUAL COSTS
C. COST EFFECTIVENESS
1. Emission reduction, Mg EO/yr (ton EO/yr)
2. Cost effectiveness, $/Mg EO ($ton EO)
Cost, $
17,100
5,130
1 855
855
32,000
9,830
65,800
9,870
• 75,700
!
i 151
76
2,480
I 70
416
812
136
3,030
', 11,900
! 19,100
0.176(0.194)
110,000(100,000)
E-36
-------
TABLE E-12. (continued)
aCalculated as 30 percent of gas/solid reactor cost.
deludes ductwork. Each aeration unit cost $16,000 installed. The example facility would require two
aeration units.
^See Section E.5 for a detailed summary of manifolding costs.
Assumed to be 15 percent of the subtotal capital costs to account for uncertainties in the capital cost
estimates.
eAssumed 15 minutes per week (52 weeks per year) for a 1,000 fP/min system and 20 minutes per week for a
3,000 ft /min control system for system inspection and general maintenance. Labor rate ($11.65 per hour)
was calculated as (323.8/218.8) x ($7.87 per hour).
'Maintenance materials were calculated as (0.5) x (maintenance labor).
SThe maximum reactant life is 1.5 years but could be shorter depending on the amount of EO being controlled
(in pounds). Assumed that control requires 4 pounds of reactant per fr/min of airflow and that each pound
of reactant can control 0.3 Ib of EO. If the pounds of EO through the control (in a 1 1/2-year period) exceed
the maximum capacity of the unit, then the reactant life is calculated as follows: [({0.3 Ib EO/lb reactant} x
{4 Ib reactant/ft3/min of flow} x {flow rate, ft3/min})/maximum capacity of the unit, in pounds of EO]*1.5
years. Reactant replacement costs were calculated as (flow rate, fr/min) x (4 Ib reactant/ft3/min of flow
rate) x ($l/lb of reactant) x (340.8/412). Where 340.8/412 is the Chemical Engineering cost indice to
convert from 1989 to 1987 dollars.hEach 1,000 ft3/min control unit requires about 8 person hours to refill.
Depending on the interval of reactant replacement, capital recovery factors were determined as [CRC = i(i +
l)n/(i + I)0"1] where i = 0.1 (10 percent interest) and n = life (years). Labor costs were calculated as
.($7.87 per hour) x (323.8/218.8) x (CRC).
'The 1,000 and 3,000 ft3/min systems are equipped with 1.1 and 10 kW (1.5 and 13.5 horsepower) fans,
respectively. Electricity costs were calculated based on 365 days per year at continuous (24 hour) operation
and an electricity cost of $0.0432 kWh. Costs were calculated as (365 days per year)(24 hours per
.day)(fan power)($0.0432 kWh).
JDisposal costs were developed assuming the reactant would be recycled and that there would be no credit or
charge for the reactant. Transportation costs were calculated assuming a distance of 1,500 miles at a charge
of $0.15 per pound for less than 5,000 pounds and $0.12 per pound for a shipment greater than
5,000 pounds.
•^Calculated as 60 percent of the sum of the maintenance labor and maintenance materials.
'Calculated as 4 percent of total capital costs.
"Assumed life of 10 years for gas/solid reactor oxidizer and aeration unit and an interest rate of 10 percent.
Assumed a life of 20 years and a 10 percent interest rate for manifolding materials. Capital recovery cost
was calculated as 0.16275/[(total capital costs)-(cost to change catalyst+labor to change catalyst+manifolding
costs)] + (0.1175) x (manifolding costs) where 0.16275 and 0.1175 are the capital recovery factors for the
catalytic oxidizer and manifolding, respectively.
E-37
-------
TABLE E-13. CAPITAL AND ANNUALIZED COSTS OF CATALYTIC OXIDATION
AT AN EXAMPLE FACILITY3
(4th Quarter 1987 Dollars)
Item
Cost, $
A. CAPITAL COSTS
1. Installed Equipment Costs
a. Catalytic oxidizer(s)
b. Installation of catalytic oxidized
c. Taxes: 5 percent of equipment cost
d. Freight: 5 percent of equipment cost
e. Aeration units, installed costb
f. Manifolding of aeration units0
g. Subtotal capital costs
fa. Contingencies^
TOTAL CAPITAL COSTS
47,800
7,170
2,390
2,390
32,000
19,700
111,000
16,700
128,000
B. ANNUAL COSTS
1. Direct Operating Costs
a. Maintenance labor6
b. Maintenance materials*
c. Catalyst replacement^
d. Labor for catalyst replacement"
e. Electricity1
f. Disposal of catalyst)
2. Indirect Operating Costs
a. Overhead"1
b. Property tax, insurance, and administration*
c. Capital recovery costs™
TOTAL ANNUAL COSTS
1,060
150
1,460
15
13,900
21
726
5,120
20,600
43,100
C. COST EFFECTIVENESS
1. Emission reduction, Mg EO/yr (ton EO/yr)(0.194)
2. Cost effectiveness, $/Mg EO ($/ton EO)
0.176(0.194)
240,000 (220,000)
E-38
-------
TABLE E-13.
/Calculated as 15 percent of catalytic oxidation cost.
Deludes ductwork. Each aeration unit cost $16,000 installed. The example facility would require two
aeration units. ' —»
°See Section E.5 for a detailed summary of manifolding costs.
Assumed to be 15 percent of the subtotal capital costs to account for uncertainties in the capital cost
estimates. r
^"S/15 minUteS ******* (365 d*yS ** yeaf) for System fc*"** *«l general maintenance of a
JlVrJ: . SyStem md **addttioaal S """tes for each step up in catalytic oxidizer size. Labor rate
($11.65 per hour) was calculated as (323.8/218.8)x($7.87/h).
Assumed $150 a year (base cost) for 1,000 ft^min unit and an additional $50 a year for each step up in
catalytic oxidizer size. F p
«The 1,000, 3,000, 6,000, 9,000, and 12,000 ft^min catalytic oxidizers have 4, 16, 32, 48, and 64 catalytic
cells, respectively. Each cell costs approximately $1,240 to refill. Costs were annualized over a 4-year life
at 10 percent interest using a capital recovery factor of 0.31547.
Each cell requires about 1-person hour to refill. Refill labor costs were annualized over 4 years assuming a
. 10 percent interest rate using a capital recovery factor of 0.31547.
'The 1,000, 3,000, 6,000, 9,000, and 12,000 f^/min systems have 80, 120, 180, 230, and 290 kilowatt
OcW) catalytic oxidizer preheaters, respectively. System designed for 70 percent heat recovery uses only
A T^ ? rated kW- Electricity cost ^culated as ($0.0432/kWh)x(kW of heater)x(0.46)x(24 hours per
day)x(365 days per year). No costs were attributed to fan electrical consumption because the preheating
.electrical costs were considerably larger.
JDisposal costs calculated as [($80/400 lb)x(90 Ib cell)x(No. of cells)]/4 years. Includes transportation to an
^ SJf c**™^ *s $25 per 55-gallon drum (7.35 fVVdnim). Density of catalyst is 1 g/cm3
(62.4 lb/ftj). Transportation equals [(90 lb/cell)x($25/drum)x(No. of cells)]/[(62.4 Ib/ft3)x(7 35 ft3/drum)
x(4 yr)]. All disposal costs were multiplied by (329.8/354.2) to correct to 1987 dollars.
^Calculated as 60 percent of the sum of the maintenance labor and maintenance materials.
'Calculated as 4 percent of total capital costs.
mAssumed life of 10 years for catalytic oxidizer and aeration unit and an interest rate of 10 percent
Assumed a life of 20 years and a 10-percent interest rate for manifolding materials. Capital recovery cost
was calculated as 0.16275x[(total capital costs)-(cost to change catalyst+ labor to change
catalyst+manifolding costs)]+(0.1175)x(manifolding costs) where 0.16275 and 0.1175 are the capital
recovery factors for the catalytic oxidizer and manifolding, respectively.
E-39
-------
E.4 EXAMPLE FACILITY CALCULATIONS
REFERENCES
1. Memorandum. Srebro, S., to D. Markwordt. EPA/CPB. Cost
Effectiveness of Reducing Ethylene Oxide Emissions from
Sterilizer Vents and Associated Vacuum Pump Drains.
March 21,'1991.
2. Memorandum. deOlloqui, V., and Srebro, S., to D. Markwordt.
EPA/CPB. Costing of Control Alternatives for the Rear
Chamber Exhaust Emissions. March 21, 1991.
I,
3. Memorandum. Srebro, S., and deOlloqui, V., to D. Markwordt.
EPA/CPB. Costing Methodology for the Control of Aeration
Room Emissions. March 21, 1991. ;
E-40
-------
E.5 MANIFOLDING COSTS
-------
TABLE E-14. INCREMENTAL CAPITAL COSTS OF MANIFOLDING
STERILIZATION CHAMBERS, 1987, $
-
1.
2.
3.
4.
OtJeninp in explosion-proof wall
Cost
Adjustable sheet metal sleeve :
Labor hours
Concrete core drilling, 1-in. to 3-in. hole
Adjustable sheet metal sleeve
Total labor hours
Labor costs at $18,91/labor hour
Overhead costs at $8.75/labor hour
Drill bole for pipe hangers
Labor hours
3/8 in. diameter hole through 1/2 in. thick steel beam, 20 holes
Labor costs at $20.32/person-hour
Overhead costs at $15.78/person-hour
Piping
Cost
100 ft, 2 in. diameter, 40 standard carbon steel pipe
90° elbows, 3 at $4.40
Tee with full-size outlet
Swing check valve
Bolts and gaskets, two sets at $7.08
Pipe hangers, 1 carton of 50 hangers
Total cost
Labor hours to install
Cut, three at 0. 16 labor hours
Bevel, three at 0.10 labor hours
Pipe
Field erection joint buttweld
Penetration through one wall
Elbows, three at 2.00 labor hours
Tee
Valve
Boltup of valve, two sets at 1.35 labor hours
Pipe hangers, 10 at 0. 16 labor hours
Total labor hours
Labor costs at $21.47/labor hour
Overhead costs at $13. 31 /labor hour
Total installed cost for piping system
Total direct costs
Total overhead costs
Administration
10 percent of total direct costs
Taxes: 5 percent of equipment costs
Total indirect costs
Total installed cost
Cost
1987, $
2a c
4.4
0.75
5.15
97a
45b
7.5
153a
118b
251ac
14ac
15ac
367ac
15ac
147ac
808
0.48
0.30
8.3
2.9
1.7
6.0
3.0
1.1
2.7
1.6
28.0
601a
373b
1,663d
536e
152f
33i
721^
2.3841
^Reference*
3-100, p. 12
3-100, p. 22
3-100, p. 22
1
3-Oi p. 2
1.0; p. 5
5-lb, p. 16
5-0, p. 1
i-o; p. 5
15-43, p. 4
1543, p.9
1543, p. 13
1543, p. 31
15-72, p. 3
15-76, p. 25
15-77, p. 27
'15-77, p. 30
15-43, p. 4
1543, p. 29
15-43, p. 8
15-43, p. 9
15-43, p. 13
15-43, p. 31
15-72, p. 3
15-76, p. 35
15-0, p. 2
1-0, p. 5
1
1-Oj p. 6
1-0, p. 6
E-42
-------
TABLE E-14. (continued)
5.
6.
Total installed cost for new vacuum pump
Total installed capital cost
Cost
1987, $
5,170)
7,554k
Reference*
MRI report2
"Direct cost.
^Overhead cost.
cEquipment cost.
''Total of costs with superscript a.
^Total of costs with superscript b.
'Administration = 0.10x(sum of costs with superscript a)
STaxes = O.QSxsum of costs with superscript c.
fTotal indirect costs = sum of costs with superscripts e, f, and g.
|Total installed cost = total direct costs (superscript d)+total indirect costs (superscript h).
J$5,000 (1986 $)x 1.034 (see Section E.2).
''Equal to total installed cost for piping (superscript i)+total installed cost for new vacuum pump (superscript j).
E-43
-------
TABLE E-15.
DUCTWORK COSTS OF MANIFOLDING CHAMBER EXHAUST
VENTS TO A SCRUBBER
Item
Cost, $,
1987
i
Reference1'3
Chamber exhaust vent to manifold3 i
1. 15 ft 10-in. diameter, 1/8 in. thick carbon steel
2. 90° elbow, 10-in. diameter
3. Laborb
210
204
164
Card, p. 4-19
Card, p. 4-22
Richardson, 15-9 p.
2
Manifold
1. 36 it 41-in. diameter, 1/8 in. thick carbon steel
2. Labor5
2,290
1,352
Card, p. 4-19
Richardson, 15-9 p.
2
Manifold to control unit
1. 30 ft 24-in. diameter, 1.8 in. thick carbon steel
2. 90° elbow, 24-in. diameter
3. Tec, 24-in. diameter
4. Laborb
1,110
580
192
769
Card, p. 4-19
Gard, p. 4-22 ;
Gard, p. 4-22 i
Richardson, 15-9 p.
2
to duct chamber exhaust vent(s) to a manifold were calculated for each sterilizer at a facility. (It was
assumed that ductwork costs for one of the sterilizers was included in the control device installation cost.)
"Labor costs developed as $3.51/fr of ductwork. ,
E-44
-------
TABLE E-16. DUCTWORK COSTS OF MANIFOLDING
AERATION UNITS TO A GAS/SOLID REACTOR
Item
Aeration room TAR) ductwork
A. AR unit to manifold3
1. 32 ft 15-in. diameter, 1/8 in. thick carbon steel
2. 90° elbow, 15-in. diameter
3. Laborb
B. AR manifold
1. 5 ft 41 -in. diameter, 1/8 in. thick carbon steel
2. Laborb
C. Manifold to control unit
1. 67 ft 24-in. diameter, 1/8 in. thick carbon steel
2. 90° elbow, 24-in. diameter costed at $580/elbow
3. Laborbc
Cost, $,
1987
710
326
481
318
187
2,480
1,160
2,140
Reference^ '^
Card, p. 4-19
Card, p. 4-22
Richardson, 15-9 p.
Card, p. 4-19
Richardson, 15-9 p.
Gard, p. 4-19
Card, p. 4-22
Richardson, 15-9 p.
2
2
2
aThese costs were developed for each aeration unit at a facility.
"Labor costs developed as SS.Sl/ft2 of ductwork.
cLabor costs include the cost to concrete core drill (24 in. hole) the aeration room wall at $784.
E-45
-------
E.5 MANIFOLDING COSTS
REFERENCES
1. Richardson Engineering Services, Inc. Process Plant
Construction Estimating Standards. 1884.
2. Beall, C., Meeting Minutes: Damas Corp. and Johnson &
Johnson. Midwest Research Institute. Raleigh, NC.
April 30, 1986. 9 p.
3. Neveril, R., Capital and Operating Costs of Selected Air
Pollution Control Systems. CARD, Inc., Niles, IL.
Publication No. EPA-450/5-80-002. December 1978.
E-46
-------
E.6 COST INDICES
-------
TABLE E-17. CHEMICAL ENGINEERING COST INDICES
Cost indices
Scrubbers
Chlorine filters
Chemicals
Operations and maintenance
labor.
Disposal of ethylene glycol
Gas/ solid reactor or
catalytic oxidizer
React ant or catalyst
replacement
Ductwork
Disposal of reactant or
catalyst
Labor for installation of
ductwork
Vacuum pumps
352.2 (1987)f*
392.1 (1989)b
352.2 (1987)°
344.1 (1987)d
340.8 (1987)®
340.0 (1986)f
323.8 (1987)9
218.8 (1988)n
323.8 (1987)1
318.4 (1986)3
352.2 (1987)*
390.7 (1989)1
340.8 (1987)m
412.0 (1989)n
323.8 (1987)°
218.8 (1978)P
329.8 (1987)3
354.2 (1989)r
323.8 (1987)f;
322.7 (1984)t
433.0 (1987)u
418.6 (1986)v
Conversion
factor
6.90
1.02
1.002
1.48
1.02
0.90
0.83
1.48
0.93
1.00
1.03
aReference 1. CE Plant Cost Index, Equipment Machinery>
Supports. October 1987 final.
^Reference 2. CE Plant Cost Index, Equipment. September 1989
final.
°Reference 1. Structural Supports and Miscellaneous.
October 1987 final.
^Reference 3. Structural Supports and Miscellaneous.
February 1986 final.
^Reference 4
fReference 3.
previous.
^Reference 1.
"Reference 1.
•^Reference 1.
^Reference 1.
^Reference 1.
Current Business Indicators.
Current Business Indicators.
October 1987 latest.
February 1986
CE Plant Cost Index, 1987 Annual Index.
CE Plant Cost Index, 1978 Annual Index.
CE Plant Cost Index, 1987 Annual Index.
CE Plant Cost Index, 1986 Annual Index.
CE Plant Cost Index, Equipment, Machinery,
Supports. October 1987 final.
•^-Reference 5. CE Plant Cost Index. Equipment, March 1989 final.
^Reference 6. Current Business Indicators, Producer Prices,
Industrial Chemicals, October 1987 (latest).
nReference 5. Current Business Indicators, Producer Prices,
Industrial Chemicals, March 1989.
°Reference 5. CE Plant Cost Index, 1987 Annual Index.
E-48
-------
TABLE E-17. (continued)
^Reference 3.
^Reference 1.
rReference 3.
^Reference 3.
^Reference 3.
uReference 4.
vReference 3.
CE Plant Cost Index, 1978 Annual Index.
CE Plant Cost Index, October 1987 final.
CE Plant Cost Index, March 1989 final.
CE Plant Cost Index, 1987 Annual Index.
CE Plant Cost Index, 1984 Annual Index.
CE Plant Cost Index, October 1987 final.
CE Plant Cost Index, February 1986 final,
E-49
-------
REFERENCES
1. Economic Indicators.
Inc., New York, NY.
2. Economic indicators.
Inc., New York, NY.
3. Economic Indicators.
Inc., New York, NY.
4. Economic Indicators.
Inc., New York, NY.
5. Economic Indicators.
Inc., New York, NY.
E.6 COST INDICES
Chemical Engineering. McGraw-Hill,
April 25, 1988. p. 9.
Chemical Engineering. McGraw-Hill,
December 1989. p. 186.
Chemical Engineering. McGrawrHill,
June 23, 1986. p. 7.
Chemical Engineering. McGraw-Hill,
December 7, 1987. p. 7. ,
Chemical Engineering. McGrawrHill,
June 1989. p. 224.
E-50
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Ethyl«ne Oxid* Emissions from Sterilization/Fumigation
Operations—Background Information for
Proposed Standards
5. REPORT DATE
March 1993
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Office of Air Quality Planning and Standards
Environmental Protection Agency
Research Triangle Park, North Carolina 27711
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
DAA for Air Quality Planning and Standards
Office of Air and Radiation
U^ S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
National emission standard to control emissions of ethylene oxide
from new and existing sterilization/fumigation operations are
being proposed under Section 112 of the Clean Air Act. This
document contains information on the background and authority,
regulatory alternatives considered, and environmental and
economic impacts of the regulatory alternatives.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
Air pollution control
Ethylene oxide
Stationary sources
c. COSATl Field/Group
Air pollution
Ethylene oxide
Pollution control
National emission standards
Industrial processes
Hazardous air pollutants
Sterilization industry
13B
8. DISTRIBUTION STATEMENT?
Unlimited
19. SECURITY CLASS (This Report/
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (Thispage}
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77)
PREVIOUS EDITION is OBSOLETE
-------
-------
-------
------- |