In recent Risk and Technology Reviews (RTR) for chemical NESHAPs, the U.S. Environmental Protection Agency (EPA) has implemented “in ethylene oxide service” requirements.

The “in ethylene oxide service” process vent requirements outlined in the Miscellaneous Organic NESHAP (MON, 40 CFR 63, Subpart FFFF) and Hazardous Organic NESHAP (HON, 40 CFR 63, Subparts F, G, and H) are of particular interest to subject facilities. The compliance date for ethylene oxide provisions for all MON affected sources has past, while HON affected sources must be in compliance upon initial startup or on July 15, 2026, whichever is later.

Both the HON and MON, by definition, assume process vents are “in ethylene oxide service” and require control if there is evidence that ethylene oxide may be present, requiring testing to prove otherwise. To demonstrate a process vent is not “in ethylene oxide service” a facility must conduct an analysis using EPA Methods 18, 320, or ASTM D6348-12 (HON only) to quantify ethylene oxide emissions are below the low threshold for “in ethylene oxide service”, set at uncontrolled emissions of 1 ppmv for each process vent or 5 lb/year from the sum of all process vents in an affected source.

For process vents designated as in “in ethylene oxide service”, emissions limits are generally much stricter (i.e. DRE 99.9%, 1 ppmv, or 5 lb/yr) than other HAP SOCMI limits. Additionally, continuous parameter monitoring systems are more stringent, with shorter 1-hour averaging periods for compliance and monitoring of additional process parameters. As a result, facilities are incentivized to complete an analysis to exempt process vents from an “in ethylene oxide service” designation.

Issues with Process Vent Testing

The “in ethylene oxide service” determination is done on a vent-by-vent basis, often making stack tests complex and resource-intensive. This is especially true for batch process vents, where testing under worst case conditions can take several days and in some cases weeks. While test methods can theoretically detect ethylene oxide levels well below the determination thresholds, specific vent conditions in the field can make achieving this accuracy challenging or practically infeasible.

Common testing complications that arise when testing for ethylene oxide can include:

• Poor Concentration Knowledge: There is often a poor engineering knowledge of trace compounds like ethylene oxide at an individual process vent level.
• High Concentration Streams: Precontrol streams have higher concentrations of analytes than the post control performance test. These streams may require dilution, which results in lower accuracy and higher detection limits.
• Improper Sampling Locations: Generally, the vent prior to and after a control device has proper sample ports. However, individual process vent headers may not be designed for EPA compliant method measurements.
• Spectral Interference (EPA Method 320): This arises when the spectral features of interfering compounds, such as water, those of the ethylene oxide, a problem amplified when testing for the relatively smaller concentrations of ethylene oxide typically present. Although, these can be mitigated by choosing alternate analytical regions knowledge of the stream composition is key.
• Temperature: A temperature that is too low may cause analytes or water vapor to condense in the sampling train. This is a particular concern as condensed moisture can remove ethylene oxide as a water soluble compound.

How to Plan for Successful Testing

As any one of the above issues has the potential to completely invalidate a stack test result and require a completely new test, significantly more planning is required for process vent “in ethylene oxide service” determinations than a typical compliance test. First ensure your selected testing firm and the onsite testing team are experienced in ethylene oxide sampling.

Second, detailed conversations with your testing firm are a must and should include a discussion of testing goals, process knowledge, sample locations, and process vent compositions to identify which of the complications listed above could be present for each process vent and strategies for how to address them.

Third, a pre-survey can be completed involving limited sampling of some or all of the process vents before the official stack test. A pre-survey will allow the source to “test” the testing strategy, identify complications, and refine the test methods and protocols to ensure the accuracy needed for the “in ethylene oxide service” determination is met. Additionally, the pre-survey can serve as an early screening tool to identify process vents likely to be classified as “in ethylene oxide service” or that may be difficult to prove are exempt from control. This can give a facility valuable time to shift focus to methods to refine accuracy or control and compliance options.

Fourth, given the combination of complex testing requirements and potential for operating delays, the test plan should allow for additional days/time beyond that for a typical stack test. When conducting a pre-survey, time should be allotted for the initial evaluation plus additional test to confirm any adjustments made. For the full compliance test, the test plan and your testing firm should have the flexibility to make adjustments to address any of the above issues as they are identified during testing. Additional days may be needed to account for delays in process operation of for evaluating initial results and conducting additional runs, as needed.

For more information and assistance evaluating your current compliance requirements, please contact Inaas Darrat, PE, at 713.552.1371 or Gena Driscoll at 251.391.5789.

Related Training:

Complimentary Webinar: HON and SOCMI Webinar Series Part 3 – New Ethylene Oxide Requirements for HON | Trinity Consultants
Effective Management of Stack Test Programs | Trinity Consultants