September 21,

2005

MEMORANDUM

Subject:

Technology

From:

David

To:

Laura

Review and Residual Risk Data Development for the Ethylene Oxide
Commercial Sterilization National Emission Standards for Hazardous Air
Pollutants (NESHAP)

Markwordt/signed/
Policy, Planning, and Standards Group
McKelvey, Group Leader
Policy, Planning, and Standards Group

This memorandum summarizes the data development effort that was conducted to
support the residual risk and technology reviews of the Ethylene Oxide Commercial Sterilization
NESHAP, as required under Section 112(0(2 ) and Section 112(d)(6), respectively, of the Clean
Air Act (CAA or Act). Section I of this memorandum presents background information on the
Ethylene Oxide Commercial Sterilization assessment is presented in a separate memorandum?
I. BACKGROUND

The NESHAP developed under section 112(d) of the CAA for the Ethylene Oxide
Commercial Sterilization source category (40 CFR part 63, subpart O) was promulgated on
December 06, 1994 (59 FR 62585). The Ethylene Oxide Commercial Sterilization NESHAP
covers ethylene oxide, the only hazardous air pollutant (HAP) emitted from the
sterilizatiordfumigation process. The emission points regulated by the Ethylene Oxide
Commercial Sterilization NESHAP are the main sterilization and aeration room vents. The
majority of sterilization facilities process on a contract basis. Some medical supply
manufacturers and spice manufacturers sterilize their own product.
The NESHAP applies to operations at both "major and area sources." A major source is
defined in CAA section 112(a)(1) 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

Memorandum, "Residual Risk Assessment for the Ethylene Oxide Commercial
Sterilization Source Category", from Mark Morris, to David Guinnup, Group Leader, Risk &
Exposure Assessment Group, OAR, US EPA, February 25, 2005. [Docket OAR-2003-0197]

 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." An area source has a
potential to emit less than a major source. During the development of the NESHAP we
estimated that there were approximately 188 facilities nationwide, of which about 47 would be
considered major. We estimated that the NESHAP would reduce emissions of ethylene oxide by

1,000

tons

The

annually.
following

is

a

summary of the NESHAP:

Main sterilization and aeration
emissions by 99 percent.

room

vents

subject

to control

requirements

must reduce

require an initial performance test, followed by continuous monitoring of
specified operating parameters, of vapor processors used to control vent emissions. The
monitoring parameter limits are to be established based on data collected during the
performance test.
The standards

II. STATUTORY AUTHORITY FOR THESE ACTIONS

Section 112 of the CAA establishes a two-stage regulatory process to address emissions
of HAP from stationary sources. In the first stage, after EPA has identified categories of sources
emitting one or more of the HAP listed in the CAA, section 112(d) calls for us to promulgate
national technology-based emission standards for sources within those categories that emit or
have the potential to emit any single HAP at a rate of 10 tons or more per year or any
combination of HAP at a rate of 25 tons or more per year (known as "major sources"), as well as
for certain "area sources" emitting less than those amounts. These technology-based national
emission standards must reflect the maximum reductions of HAP achievable (after considering
cost, energy requirements, and non-air health and environmental impacts) and are commonly
referred to as maximum achievable control technology (MACT) standards.
For area sources, CAA Section 112(d)(5) provides that in lieu of MACT, the
Administrator may elect to promulgate standards or requirements which provide for the use of
generally available control technologies or management practices and such standards, are
commonly referred to as generally available control technology (GACT) standards. GACT
standards are also technology based standards.

MACT for major sources under section 112(d)(2). As for
sources, we established a MACT standard for certain emission points pursuant to section
112(d)(2), and a GACT standard for other emission points pursuant to section 112(d)(5).
In that final

Section

rule,

we

set

112(d)(6)provides

that EPA review these

technology-based

area

standards and revise

 them "as necessary

technologies)"

no

(taking into account developments
less frequently than every 8 years.

in

practices,

processes and control

The second stage in standard setting is described in section 112(f) of the CAA. This
provision requires, first, that EPA prepare a Report to Congress discussing (among other things)
methods of calculating risk posed (or potentially posed) by sources after implementation of the
MACT standards, the public health significance of those risks, the means and costs of controlling
them, actual health effects to persons in proximity to emitting sources, and recommendations as
to legislation regarding such remaining risk. The EPA prepared and submitted this report
("Residual Risk Report to Congress," EPA-453/R-99-001) in March 1999. The Congress did not
act on any of the recommendations in the report, triggering the second stage of the standardsetting process, the residual risk phase.
Section 112(f)(2) requires us to determine for each section 112(d) source category
whether the national emission standards protect public health with an ample margin of safety. If
the national emission standards for HAP "classified as a known, probable, or possible human
carcinogen do not reduce lifetime excess cancer risks to the individual most exposed to
emissions fi-om a source in the category or subcategory to less than one in one million," EPA
must promulgate residual risk standards for the source category (or subcategory) as necessary to
provide an ample margin of safety. The EPA must also adopt more stringent standards to prevent
an adverse environmental effect (defined in section 112(a)(7) as "any significant and widespread
adverse effect.., to wildlife, aquatic life, or natural resources
"), but must consider cost,
energy, safety, and other relevant factors in doing so.
Section 112(f)(5) expressly provides, however, that EPA is not required to conduct any
review under section 112(f) or promulgate any emissions limitations under that subsection for
any area source listed pursuant to section 112(c)(3) for which EPA has issued GACT standards.
Thus, although EPA has discretion to conduct a residual risk review under section 112(f) for area
sources for which it has established GACT, it is not required to do so. See CAA section
11
III. DATA DEVELOPMENT FOR PERFORMING THE RESIDUAL RISK REVIEW
Our approach for conducting the data development effort necessary to support the
residual risk assessment for the Ethylene Oxide Commercial Sterilization NESHAP included two
primary objectives. The first objective was to gather the data necessary to determine the existing
residual risk from facilities in the source category that are subject to the control requirements of
the NESHAP. The second objective was to determine the potential for achieving further
reduction of emissions (and, thus, the risk) from those facilities. Section III-A of this
memorandum discuss the development of the data needed to perform the residual risk assessment
and Section II/-B discusses our evaluation of additional controls available to reduce the risk.
A.

DATA GATHERING

 To evaluate the residual risk from the Ethylene Oxide Commercial Sterilization source
category after implementation of the Ethylene Oxide Commercial Sterilization and Fumigation
NESHAP we identified those facilities subject to the control requirements of the standard and
described their site locations, operations, and emissions. During the development of the
Ethylene Oxide Commercial Sterilization and Fumigation NESHAP, EPA gathered information
from a variety of sources to describe and characterize the industry, the amount of HAP emitted,
the emission sources, the control measures that were in use or available, and the projected
emission reductions and costs of implementing the standards. However, the data were primarily
gathered prior to 1987 and significant consolidation in the industry made it necessary to reassess
plant emission characteristics. The following paragraphs describe the data gathering activities
conducted for this residual risk study. Approximately 100 facilities were ultimately identified to
be commercial sterilizers.

Source Identification. b The data were gathered primarily from the following three
sources: (1) the 1999 National Emissions Inventory (NEI); (2) the 2000 Toxics Release Inventory
(TRI); and (3) The Ethylene Oxide Sterilization Association (EOSA) review of the data set. A
limited number of facilities were contacted directly to verify the data.
In the U.S. 74 ethylene oxide sterilization facilities were identified from the NEI data
base. The 1999 NEI primarily provided location, emission point, control efficiency, and stack
characteristic data, though the latter two types of data were available for only about a third of the
ethylene oxide facilities identified in the NEI.

The 2000 TRI became available in June 2002. We subsequently extracted facility
identification, location, emissions, and control efficiency data for all ethylene oxide emitters (157
facilities), reduced the list to ethylene oxide sterilization facilities by eliminating obvious
chemical manufacturers and other non-ethylene oxide facilities, and refined and corrected the
data as needed (e.g., for facilities that did not report fugitive emissions). We made several phone
calls and conducted interact searches to determine whether some of the larger facilities were still
operating and, if so, what their status was in terms of compliance with the MACT.
TRI data from 1999 and earlier data were used for some facilities if 2000 data were not
available or we needed to assess changes in emissions over time. TRI data also were useful for
determining ethylene oxide use.
we identified as being subject to the Ethylene Oxide Commercial
Sterilization and Fumigation NESHAP cover a range of processes, throughputs, emission levels,
product mix, and locations.

The facilities

b Memorandum, "Data and assumptions used for the screening-level residual analysis of
the commercial ethylene oxide sterilizers and fumigators source category", from Alex Koppel,
Jim Laurenson, Marsha Fisher and David Burch, ICR to Mark Morris and David Markwordt
OAR, US EPA, June 09, 2004. [Docket OAR-2003-0197]

4

 Source Characterization To assess th• risk posed by HAP emissions from the ethylene
oxide commercial sterilization source category, detailed site specific information is necessary.
Modeling inputs such as latitude and longitude of the emission points, HAP emission, rates, stack
parameters, and distance to fence line are used to perform the actual risk modeling. Other
variables such as information on the number and capacities of emission sources, and knowledge
of the types and efficiencies of control measures in use are used to help understand the modeling
results.
Not all the data fields in the NEI are populated with data; and while the entries for some
facilities had details on each emission source, many others had partial data or data for the total
facility only. For missing key parameters (such as stack height, stack diameter, and exit gas
temperature and velocity) the NEI uses default entries based on industry average values. Latitude
and longitude defaults are based on the facility street address and ZIP code.

however,

supplement
industry representatives.

the NEI data base with information
obtained from industry contacts or
The Ethylene Oxide Sterilization
Association (EOSA) submitted detailed data several times over the course of this data gathering.
In August 2002, EOSA updated or provided data for all modeling parame.ters for approximately
20 facilities. In March 2003, EOSA submitted updated data for both the August 2002 facilities
and an additional 26 facilities for a total of 46 facilities.
In

some

cases,

we were

able to

The residual risk modeling identified 44 of 76 facilities with a census block individual
lifetime cancer risk at or above 1 in a million, 19 with a census block individual lifetime cancer
risk at or above 10 in a million, and 0 with a census block individual lifetime cancer risk at or
above 100 in a million,
B. EVALUATION OF ADDITIONAL CONTROLS AVAILABLE TO REDUCE RISK

Facilities in the Ethylene Oxide Commercial Sterilization source category typically emit
oxide from the sterilization chamber and the aeration room. There are 3 emission vents
associated with the sterilization and fumigation process: the main sterilization and the chamber
exhaust vents from the sterilization chamber and the aeration room vent from the aeration room.
As explained below the chamber exhaust vent is not subject to control. The current standard
requires MACT control of emissions from the main sterilization vents at both major and area
sources. The standards also require MACT control of emissions from aeration room vents from
major sources. This section will present the technological feasibility of additional controls to
reduce residual risk from this emission source category.

ethylene

Sterilization chambers are filled with products, sealed, and infused with ethylene oxide to
sterilize/fumigate the product. At the conclusion of the sterilization period, the ethylene oxide is
pumped from the chamber to a control device. After sufficient removal of ethylene oxide from
the chamber, the sterilized product is removed and placed in an aeration room. The purpose of
the aeration room is to drive off any ethylene oxide residual left in the product after removal of

 ethylene

oxide from the sterilization chamber.
routed to a control device.

Ethylene

oxide driven off in the aeration

room

is

Many, if not all, source facilities utilize a chamber exhaust fan while personnel are
removing product from the sterilization chamber. This fan removes ethylene oxide off-gassing
from the product. The Ethylene Oxide Commercial Sterilization and Fumigation NESHAP
promulgated in 1994 (59 FR 62585) required control of the chamber exhaust vent. In 1997 there
were a series of explosions associated with control of the chamber exhaust vent (62 FR 64736).
We subsequently reassessed the control requirements and removed the requirement to control the
chamber exhaust in November 2001 (66 FR 55577); the Agency continues to believe that the
action taken in 2001 is reasonable and we have found no safe way to impose controls on the
chamber exhaust vents, Approximately 1 percent of the ethylene oxide used in the process is
emitted through the chamber exhaust vent.
The Ethylene Oxide Commercial Sterilization and Fumigation NESHAP requires that a
control device must achieve a 99 percent reduction in ethylene oxide emissions for both the main
sterilizer vent and aeration room vent subject to emission reduction requirements. Current
control technology used to control ethylene oxide emissions includes scrubbers, combusters, drybed absorbers, or a combination of these devices. During our investigation of the safety issue
associated with the chamber exhaust vent we did not find any new technology for the any of the
vents. Some facilities are already using combinations of the current control technologies to
achieve the standards. We have no data supporting a further reduction in emissions by adding
down-stream control devices to existing control equipment achieving the standards. Major and
area source vent characteristics are similar so the same technologies are used for both major and
area sources. Therefore, our conclusions concerning new technology apply to the all vents.

Ethylene

by using alternatives to ethylene oxide
plasma,
hydrogen
peroxide,
and ozone). However, these chemicals
gas
do not necessarily offer environmental improvements over ethylene oxide. None of these
alternatives can replace the use of ethylene oxide in all applications. More importantly, ethylene
oxide is extremely effective as a sterilant.(see appendix).

(e.g.,

chlorine

oxide emissions could be eliminated

dioxide,

The Food and Drug Administration (FDA) has primary authority to regulate the use of
sterilization methods. The FDA issued guidance [510(k) Sterility Review Guidance K90-1,
August 30, 2002 ("FDA Guidance")] to facilitate non-traditional sterilization methods. The FDA
stated in the guidance that the FDA "has had little or no experience with these methods for
achieving sterilization and is concerned about a manufacturer's ability to successfully use such
methods without adversely affecting the sterility assurance level...". If the use of ethylene oxide

release, Sterigenics, Incident in Sterigenic's Ontario, California Facility, August
20, 2004: "Sterigenics Intemational, Inc. announced today that it had a serious incident involving
an explosion within its Ontario, California facility on August 19, 2004." California currently
requires control of rear chamber exhaust. [Docket OAR-2003-0197]
Press

 prohibited, manufacturers of products requiring sterilization will have to reconsider the
device and packaging material, its compatibility with the nontraditional sterilizing agent, the
packaging configuration,, the ability of the nontraditional sterilant to penetrate the packaging,
cost, and availability. Because these nontraditional sterilization methods are less known,
were

the

manufacturers would have to submit to FDA their validation data for review. Nontraditional
sterilization operations can not be used to sterilize materials until they have been validated.
Prohibiting the use of ethylene oxide carries the risk of creating a void where some products may
not be able to be sterilized until newer systems are designed and validated. Until such time as
these non-traditional sterilization techniques may be used under FDA. rules, these techniques are
not considered available for the purpose of reducing emissions.
Radiation (gamma and electron beam) can be used to sterilize many products. Radiation
sterilization has been used for about half of the products sterilized in the United States. However,
all sterilization techniques are limited in their applications. For example, gamma radiation has
potentially damaging effects on products, particularly those products that contain polymers. And,
radiation technology is completely different from chamber sterilization. "Ethylene oxide and
radiation technologies (both gamma and e-beam) share no common equipment. Any conversion
would involve scrapping the. ethylene oxide chambers and the related specialized equipment and
systems and likely displace most if not all the existing workforce, in effect shuttering the ethylene
oxide sterilization industry. Additionally, not even the buildings could be saved. To construct a
radiation facility special shielding (huge concrete/lead shields) and storage pools need to be
incorporated in to the design of both the building and the process. ''d
EPA did not quantify the costs associated with an ethylene oxide prohibition. We believe
the annual costs of a prohibition would likely exceed the annul costs of the original control
requirements i.e., approximately $7,000,000 per year.

III. TECHNOLOGY REVIEW
In addition to the requirements in section 112(0(2) to review the residual risks, section
(d)(6) of the CAA requires us to review, and revise as necessary (taking into account
developments in practices, processes, and control technologies), emission standards promulgated
under this section no less often than every 8 years. The Ethylene Oxide Commercial Sterilization
and Fumigation NESHAP was promulgated on December 6, 1994. We subsequently reassessed
the MACT requirements and removed the requirement to control the chamber exhaust in
November 2001 (66 FR 55577). Our investigation did not identify any significant developments
in practices, processes, or control technologies since promulgation of the original standard in
112

1994.
Letter from Joseph E.
facilities could be retrofitted to
d

0197]

Hadley, Jr.,
use

EOSA to David Markwordt, EPA, "Whether EtO
radiation technology," October 5, 2004. [Docket OAR-2003-

 Existing

Sources:

conclude there are no significant developments in practices, processes, or
control technologies for both existing major and area sources, the rationale for both controlled and
uncontrolled vents in the original MACT and GACT determinations remains unchanged. [59 FR
10591, March 7, 1994 and FR 59 62585 December 6, 1994]
Because

New

Major

we

Sources:

While no additional control technologies have been identified that could provide improved
control of emissions, we are aware of existing State rules which have control limits exceeding the
99% MACT requirement. The State of California emission reduction requirement for the main
sterilizer vent is 99.9% reduction; this requirement was enacted prior to promulgation of the
Federal requirements.
We do not have data to confirm that all facilities are capable of achieving 99.9% on a
continuous basis. Prior to 1994, we did limited testing; the tests were performed to develop a
sampling train necessary for performance testing of ethylene oxide control technology. The
results of one of the three tests performed showed the control unit with an emission reduction
slightly below 99%. In 1994, it was our understanding California did not have sufficient
performance tests to demonstrate continuous compliance with 99.9% for all operating conditions
for both scrubbers and combustion units within the industry. Therefore, in 1994, in support of the
Federal control limit, we concluded both rules are sufficiently stringent to require application of
the same technologies. We believe it reasonable to assume the same technologies perform
similarly i.e., those facilities outside of California are on average likely to achieve emission
reductions similar to those inside California.
Both the EPA and California rules require a test to demonstrate compliance with the
emission reduction limit and continuous monitoring to ensure proper operation and maintenance.
We believe these test requirements are sufficient to ensure maximum achievable control
technology is in place and operated properly. Initial compliance tests are performed one time and
on a very narrow set of operating conditions. These test results are too limited to determine if
there are any meaningful differences in control technology lifetime performance associated with a
99% and 99.9% performance limit.
In the absence of data

requirements

for

new

major

supporting a higher
facilities will remain

control limit for

unchanged.

new

major

sources, the control

New Area Sources:
In the
because of the

original area source determination, MACT for new area sources was rejected
high incremental cost effectiveness. The Administrator employed his authority

 under section 112(f) of the Clean Air Act to base the standards for new area sources on GACT.
We have no reason to believe the capital cost developed in the original determination have
decreased. Therefore, the rationale for GACT for new area sources remains unchanged. [59 FR
10591, March 7, 1994 and FR 59 62585 December 6, 1994]

 Appendix

Alternative Sterilization Process

Papers

M.J., DeGragne P., Olson N., Puchalski T. Comparison of Ion Plasma, Vaporized
Hydrogen Peroxide, and 100% Ethylene Oxide Sterilizers to the 12/88 Ethylene Oxide Gas
Sterilizer. Infection Control and Hospital Epidemiology, February 1996; 17(2): 87-91.
1. Alfa

Calhoun L.R., Allen J. T., Shaffer H.L., Sullivan G.M., Williams C.B. Electron-Beam
Systems for Medical Device Sterilization. Medical Plastics and Biomaterials Magazine, July

2.

1997.
Alfa M.J., DeGragne P., Olson N. Bacterial Killing Ability of 10% Ethylene Oxide Plus 90%
Hydochlorofluorocarbon Sterilizing Gas. Infection Control and Hospital Epidemiology, February
1996; 18(9): 641-645.
3.

Feldman L.A., Hui
Temperature Hydrogen

4.

H.K. Compatibility of Medical Devices and Materials with LowPeroxide Gas Plasma. Medical Device & Diagnostic Industry Magazine,

December 1997.
5.

Arscott

Parente D. Validating Radiation Sterilization
Diagnostic Industry Magazine, February 1999.

E., Anderson S., Broad J.J.,

Marketplace.

Medical Device &

6. Hemmerich K.J.,
Device & Diagnostic

Polymer Materials Selection for Radiation-Sterilized
Industry Magazine, February 2000.

Nighswonger G., Keeping a Weather Eye
Industry Magazine, February 2001.
7.

on

Mosley G.A., Gillis J.R.,
Determining the Lethality of
Magazine, February 2002.
9.

Using Dense-Phase
Magazine, May 2001.

Approaches
Technology,

11. Allen D. Sterilization Combination Products. Pharmaceutical & Medical
October 2004.

10

Global

Diagnostic

Whitbourne J.E. Calculating Equivalent Time for Use in
EtO Sterilization Processes. Medical Device & Diagnostic

10. Allison D.G. and Mohammad S. Traditional or Non-traditional
Sterilisation of Medical Devices. Medical Device Manufacturing &

a

Products. Medical

EtO Sterilization. Medical Device &

8. Matthews M.A., Warner L.S., Kaiser H. Exploring the Feasibility of
Carbon Dioxide for Sterilization. Medical Device & Diagnostic Industry

in

Industry

to the

June 2002.

Packaging News,