Date:		2 July 2019
To: 		Jonathan Witt, U.S Environmental Protection Agency
From:		Meaghan McGrath, RTI International
		Karen Schaffner
Contact:	Greg Crist
		AdvaMed
		https://www.advamed.org/

 Participants

AdvaMed

Greg Crist, Chief Advocacy Officer and Head of External Affairs, AdvaMed
Janet Trunzo, Senior Executive Vice President, Technology and Regulatory Affairs, AdvaMed
Ellen Kondracki, Vice President, Sustainability & EHS, BD 
Dan Carestio, Chief Operating Officer, STERIS
Whitney Tull, Senior Director, Government Affair, STERIS
Dianne Gibbs, Regulatory Affairs Manager, 3M Medical Solutions Division

Anguil Environmental Systems

Chris Anguil, President

U.S. Environmental Protection Agency

Jonathan Witt, Fuels and Incineration Group
Ned Shappley
Margaret Sieffert, Region 5
Tess Petesch

RTI International

Meaghan McGrath
Karen Schaffner

 Discussion
On July 2, 2019, representatives of the U.S. Environmental Protection Agency (EPA) and their contractor (RTI International) participated in a meeting with representatives of AdvaMed.  Mr. Greg Crist began the discussion by providing a brief background on AdvaMed and its association with the ethylene oxide (EtO) sterilization industry. AdvaMed is a trade association that focuses on advancing medical technology.  AdvaMed is comprised of 400 member companies, many of which are medical companies (for example, Medtronic, Bayer, STERIS, BD, etc.).  Approximately two-thirds of medical products are sterilized using EtO.
Discussions Led by BD
Ms. Ellen Kondracki provided background information on Becton Dickinson (BD), as well as BD's history within the EtO sterilization industry. Ms. Kondracki has worked in the chemical industry for 15 years, focusing mainly on regulatory compliance and has chaired AdvaMed's sterilization workgroup for the past few months. The sterilization workgroup is made up of approximately 70 members that are subject matter experts (SMEs) or manufacturers that use EtO in sterilization operations. The sterilization workgroup has been closely following the news on certain sterilization facilities that are receiving intense public scrutiny. Ms. Kondracki noted that the workgroup has focused discussions lately on the viability of alternatives to EtO sterilization, as well as focusing on the environmental health and safety effects and controls used within the industry to promote safety in the workplace and to reduce air emissions.  With respect to alternatives to EtO, one of the biggest barriers to finding a viable alternative to EtO is scalability of alternatives.
BD has approximately 80 manufacturing facilities worldwide. Of these 80 plants, there are many that conduct in-house sterilization.  BD has four sterilization facilities in the U.S. that use EtO, as well as others using gamma radiation or other forms of sterilization. Some of the BD EtO facilities are larger-scale sterilizers and some are smaller-scale sterilizers and use of sterilizers with separate aeration units is typical for their facilities.  With respect to air pollution control devices (APCD), BD EtO facilities exclusively use thermal or catalytic oxidizers; two sites use catalytic oxidizers, and two sites use thermal oxidizers. Their facilities do not operate wet scrubbers. Across these facilities, the emission sources that are controlled varies. For larger sterilizers, the chamber exhaust vents (CEV) are controlled with catalytic oxidizers, and at smaller sterilizer facilities, work practices such as increased gas washes are conducted. BD is currently working with one or two vendors to improve operation of their catalytic oxidizers.
Sterilization Cycles that Use Less EtO
BD is focused on designing sterilization cycles to use less EtO.  They have optimized and changed their sterilization cycles to use less EtO and can achieve at least a 33 percent reduction in EtO usage on average at their facilities through this reduction strategy.  BD is also working on a process with automated movement of the product from the sterilizer chamber to the aeration unit.
The steps to revise sterilization cycles requires re-validation with the U.S. Food and Drug Administration (FDA).  The sterilization cycle depends on the classification of the medical device (for example, Class I, II, or III devices, lowest risk to highest risk devices).  Based on the class of device, some sterilization cycles are easier to revise than others. It takes a good bit of work from a quality standpoint to redo validations. To comply with FDA requirements, manufacturers must have a quality management system in place that dictates how a change occurs to a sterilization cycle. For EtO sterilization cycle revisions, the FDA would review the data to make sure validation requirements are met. For small manufacturers, it is difficult to know how cost-prohibitive revising the sterilization cycle might be. EtO sterilizations tends to sterilize higher classes of devices, for example, complex or delicate medical equipment.
BD built their sterilization facilities in the U.S. in approximately 2000.  Most of the air pollution controls were designed to control the CEV, i.e., there has been no retrofitting for controls of CEVs.  Process engineers at BD have been involved with the sterilization industry from the 1980's and 1990's and were aware of the need to accurately vent to air pollution controls, as is especially important for EtO facilities.  BD shared some estimates for the cost of tying in the CEV.  The cost of ductwork, dampers, etc. for routing the CEV to a control device depends on the situation, but cost is generally in the range of $0.5 million to $1.25 million for tying the emission source into existing air pollution controls. Control of the CEV would also require use of a snubber in line prior to the oxidizer APCD to moderate concentration.
For the four U.S. BD facilities, the EtO concentration is monitored in the sterilizer chamber during gas washes. Specifically, continuous in-chamber EtO concentration monitoring is conducted at the two larger U.S. sterilizers.  BD was not certain of the measurement technique used.
Prior to initiating the CEV on the sterilizer chamber, a certain percent (%) of the lower explosive limit (LEL) or EtO concentration in the sterilizer chamber must be reached.  In addition, interlock systems are used on the sterilizer chamber door that require the EtO concentration be reduced to a certain level prior to opening.  The EtO concentration is monitored within the chamber, and this concentration is linked to the door interlock. The concentration must be less than or equal to 1 milligrams per liter (mg/L).
Smaller sterilizers emit approximately 50 pounds per year (lb/yr) of EtO from the stack. Larger sterilizers emit approximately 100 lb/yr. BD uses separate aeration units at all of their facilities.  Some of their facilities are automated, and in the future, they intend for all of their plants to be automated.
For the dwell period in the sterilizer chamber, the EtO concentration has historically been in the range of 600 mg/L, but this concentration has been declining over the last decade, and this concentration is now closer to approximately 400 mg/L.
Removal efficiencies for EtO typically range from 99.95 percent at the low end to 99.99 percent.  The destruction efficiency for a thermal oxidizer depends on the temperature maintained. 
The aeration rooms have continuous monitoring and are vented to oxidizers. To ensure EtO is routed to a control device and is not being emitted as fugitive emission, EtO concentration monitors measure within the aeration room.
Recap from AdvaMed Meeting with B. Wehrum 
AdvaMed met with Mr. Bill Wehrum, the EPA Assistant Air Administrator, this summer.  AdvaMed wanted to discuss fugitive emissions and how those will be considered in the future, and alternatives to EtO sterilization. Mr. Wehrum asked multiple questions about sterilization process operations, including maintaining negative pressure in room areas and issues with venting fugitives from room areas at EtO facilities to APCDs.  AdvaMed noted that the industry has long been pursuing alternatives to EtO, however the medical product industry is still reliant on EtO sterilization. A number of EtO alternatives work at the prototype level, but none of the alternatives are scalable to the size needed to replace EtO in the industry. 
Discussions Led by STERIS
STERIS has nine large EtO sterilization facilities that are controlled, and the company's international facilities also meet the U.S. standards.  STERIS facilities for the most part sterilize medical devices, and approximately one-third use EtO. STERIS is a technology neutral company and is focused on using the best sterilization technology for the customer/product, whether that technology or approach uses EtO or another type of sterilization process. STERIS utilizes many different types of abatement and air pollution controls for EtO at their facilities. Two years ago, STERIS launched a Sustainable EO program, and their goal is to reduce EtO usage/inputs by 50 percent in 5 years across their organization at a global level.
All EtO sterilization operations have the same basic configuration, that includes a sterilizer vessel, aeration rooms, and EtO drum storage rooms.  At STERIS facilities, all of the emission sources are vented to control devices, including sterilizer chambers, CEV, and aeration.  Controls may vary depending on the size of the facility.  Smaller sterilizers may vent to dry bed scrubbers, while larger sterilizers may use Glygen systems, LESNI's balancer/abator system, and catalytic oxidizers as well. 
APCDs at STERIS facilities typically achieve greater than 99 percent reduction of EtO.  In California, STERIS has two large facilities with controls that achieve 99.99 percent removal efficiency, and in Texas, a facility that achieves a 99.5 percent removal efficiency of EtO. 
For the chamber exhaust, STERIS has retrofitted to route the CEV to existing APCDs.  It depends on the facility, but costs are likely to be $0.5 to $1.25 million. In 2008 when other facilities had major issues with explosions related to the CEV, STERIS reconfigured their facilities to control all CEV and to vent all CEV flow through wet "bubbler" systems (i.e. snubbers) prior to the APCD.  While their facilities do not run in the flammability zone, and there is no concern that the gas will reach the flashpoint, the wet system scrubber provides an added layer of safety precaution. The CEV emissions are routed to a wet bubbler system so that if a `flashback' occurs, it will not reach back to the sterilizer chamber.
Following aeration operations, sterilized product is moved to the warehouse.  STERIS facilities store products for customers for a minimum amount of time because of capacity constraints, and customers pick up their materials as soon as possible after sterilization. In the warehouse, sterilized products may be staged in outbound for approximately half a day until the customer's truck arrives for loading. STERIS facilities do not have space to warehouse products. 
At some STERIS facilities, exhaust fans in the warehouse storage area handle the air and venting.  Baseline EtO monitoring is conducted in the warehouse room area, and employees wear personal protective equipment (PPE).  
The aeration rooms are under engineering controls and are balanced to ensure negative pressure is maintained in the room. The facilities use baseline monitoring at leak points (such as door openings) to ensure EtO is not escaping the room. STERIS has installed Rytec doors, that are considered "zip tight" doors, in all of their facilities, which means that there is little time with the door open even when adding and removing product from the aeration room.
STERIS facilities in El Paso, TX, Spartanburg, SC, and Northborough, MA (near Boston) each have a similar design and follow typical sterilization process setup. The STERIS facilities in Grand Prairie, TX (near Dallas) and San Diego, CA use a combination-type sterilizer chamber, which STERIS refers to as "EO Express." In EO Express, all sterilization operations are conducted in all-in-one processors where preconditioning, sterilization, and aeration all are conducted in one chamber (i.e., in the sterilization chamber itself). Because these additional activities happen in the sterilizer chamber, the residence time for products in the chamber is 20 hours instead of the normal 10. STERIS is the only company that is able to operate combination sterilizers on a large, contract scale in the U.S. The pressure vacuum gradients that this process requires are not appropriate for all types of products, for example, surgical kits could not go through this process.
Use of combination sterilizers eliminates the CEV.  The combination-type sterilizer chambers tend to have hood vents over the sterilizer chamber door so that any fugitive EtO that vents from the door at the front of the chamber is captured; there are also sweep vents in the floor. Generally speaking, the EtO Express process is a cleaner process, since aeration is conducted within the chamber and there are no CEV emissions. 
At STERIS facilities, most of the hoods over combination sterilizer chambers are routed to a control device, and if not, they will be routed to a control device in the near future. At some sites, the hoods over combination sterilizers are not yet tied in to the APCD but this is fairly easy to do (simply a function of installing ductwork and the time to design and implement).  These fugitive emissions consist of high air flowrates and low EtO concentration.
The ventilation air in the sterilizer chamber room area, where the sterilizer vessels are operating, is vented to a control device. The aeration room (i.e., this is referring to the emission source, not fugitives) is vented to a control device as well.
Some members of AdvaMed's EtO workgroup have in-house sterilization, others use contract sterilizers to sterilize.
Discussion Led by Anguil Environmental Systems
Anguil Environmental Systems (Anguil) is a manufacturer of thermal and catalytic oxidizers for EtO sterilization and other industries. Anguil has approximately 1,900 installations worldwide and has been providing EtO abatement since the 1990's. Anguil guarantees destruction efficiency of 99 percent plus for EtO. For high inlet concentrations of EtO, 99 percent or 1 part per million, volume (ppmv) is easy to achieve, but for lower concentration streams (e.g., aeration rooms), there are concerns regarding the outlet testing methodology (EPA Method 25A, gas chromatography, etc.) used to measure EtO destruction because there is some inaccuracy/play in the results. Anguil can guarantee 99.9 percent for high EtO concentration streams, but it is hard for a manufacturer to guarantee 1 ppmv if the inlet concentration to APCD is less than 20 ppmv. 
EPA noted that EPA Methods 18 and 25A can measure EtO concentrations less than 100 ppmv, and the detection limit is understood to be approximately 0.3 ppmv.  EPA noted that a number of monitoring manufacturers are citing monitoring equipment that is more sensitive for EtO, with detection limits of approximately 1 to 20 parts per billion, volume (ppbv). Optimized Fourier Transform Infrared (FTIR), using EPA Method 320, is sensitive and has a detection limit of approximately 1 ppbv.  EPA would need to review associated test data to demonstrate the performance of the monitors.  It is anticipated that some of these newer monitoring devices will be commercially available in roughly 6-month timeframe.  Anguil also noted they are aware of the newer approaches in monitoring but do not have the data.
The Anguil oxidizers can handle emissions from sterilizer chamber vents (SCV), CEVs, and aeration room vents (ARV) depending on the facility. At older locations, it is quite common to have CEVs vented to the oxidizers. Facilities that have a peak shaver/quench snubber installed following the SCV and prior to the APCD are able to vent to an oxidizer.  High concentration flows from the sterilization chamber are directed to a water tank (i.e., peak shaver/quench snubber) where the EtO is bubbled through and absorbed in the water.  Then the EtO is stripped from the water at a constant rate and is mixed with the ARV air prior to the oxidizer. 
The test method to determine the destruction efficiency (DE) of the oxidizer is selected by the end user, and a third party conducts the emission testing. Anguil noted they are not sure if a capture test was conducted in determining DE provided above.
The size of the APCD used at a facility is based on the air flow from the processes. Anguil has built small catalytic oxidizers from 100 standard cubic feet per minute (scfm) and larger ones up to 20,000 scfm. The largest catalytic oxidizer available right now is 20,000 scfm and is designed for ARVs only.
Anguil can likely share some capital and annual cost data with EPA but noted that cost really depends on the scope of the control system, as there are many variables that affect price. A higher flow system that uses more natural gas means increased capital and annual cost. For capital costs for an APCD that include capture improvements to achieve a combined capture and control of 99.9 percent, a back-of-the-envelope calculation for capital costs could be $100 million. Smaller systems could range from a few hundred thousand dollars to a couple million dollars. Hospital sterilizers could be hundreds of dollars per cfm.
Anguil has oxidizer control devices for smaller sterilizer units in California.  These sterilizer units have typical EtO usage of 25 lb/yr.  It is sometimes difficult for Anguil to provide a performance guarantee when it is not clear which test methods would be used by a state to determine the outlet EtO concentration in ppmv.
With respect to the catalyst bed replacement schedule, facilities often have their preferred approach. Some facilities replace part of catalyst beds, for example, replace one-third of the catalyst every year, and do not do a complete replacement of the catalyst at one time. Other facilities do a full replacement of the catalyst. Facilities can conduct testing on the activity of the catalyst and compare to the specification of a new batch of catalyst by measuring via x-ray or gas chromatography.  Facilities would take a sample plug of catalyst or a small core of catalyst and send to a laboratory, and often send the catalyst sample to the catalyst manufacturer for testing.  Anguil was not aware of an American Society for Testing and Materials (ASTM) method for testing the activity of catalyst. Over time, the catalyst may have a slight decrease in performance, however a small increase in the operating temperature of the catalyst bed will improve efficiency/performance; there is a tradeoff in operating costs at higher temperatures versus replacing the catalyst. In general, the catalyst life averages 5 years. Anguil offers replacement services as part of their contracts. 
Anguil noted that some facilities monitor outlet EtO concentrations from the oxidizers using continuous emission monitoring systems (CEMS), but this is third party monitoring equipment.  Facilities determine the operating temperature based on the temperature during the stack test performance demonstration, and this sets the operational specification for temperature (i.e., operate the oxidizer at the temperature during the stack test).
In general, the stack test is conducted at worse case.  This is often defined as high EtO concentrations, but some industries or states may require demonstration of the emission standard at both high and low EtO concentrations.  Performance for processes with low inlet concentration are more difficult to guarantee.
Controls for Fugitive Emissions from EtO Sterilization Facilities
There has been a good bit of discussion in the industry regarding control of fugitive emissions. In the industry as a whole, there is significant variation in facility age and design. The variation in facility layout often comes from whether the facility is a commercial sterilizer or a manufacturing facility that also happens to conduct EtO sterilization. In the past, fugitive emissions have been a focus of the EtO environmental health and safety (EHS) subgroup from a worker safety or employee exposure perspective (e.g., OSHA PEL).  Today, the focus is about benchmarking what emissions occur and how emissions are currently controlled. A number of companies have been very public about improvements they are making at their facilities with respect to fugitive emissions. There are many workgroup participants who are capturing EtO emissions at the sterilizer chamber.  After the sterilizer chamber, EtO emissions occur from off-gassing from the sterilized product in the room areas, in aeration process, and from the warehouse area.  Some warehouse areas are climate-controlled in the winter as the area gets heated for operator comfort, however in the summer the room can be hot (ambient temperature).  Drum rooms, etc. could also be vented to a control device. Decreasing the initial EtO usage is also a solution. 
Typically, at AdvaMed facilities there is a building management system where facility operations managers are watching concentrations, etc. in the different steps of the EtO sterilization process.
Closing
STERIS welcomed EPA to visit their Spartanburg, SC facility and possibly their Grand Prairie, TX facility. There are a number of validation/technical staff at the Texas facility who are able to discuss measurement specifics of EtO operations.
