Monitored and Modeled Ambient Air Concentrations of Ethylene Oxide: Contextualizing Health Risk for Potentially Exposed Populations in Georgia

Abstract

Recent studies have monitored and modeled long-term ambient air concentrations of ethylene oxide (EO) around emitting facilities in Georgia with the intent of informing risk management of potentially exposed nearby residential populations. Providing health context for these data is challenging because the U.S. Environmental Protection Agency's risk-specific concentrations lack practical utility in distinguishing a health significant increase in exposure. This study analyzes EO data for eight emitting facilities, using a previously published alternative exposure metric, the total equivalent concentration, which is based on U.S. Centers for Disease Control biomarker data for the non-smoking U.S.

Population: Mean concentrations for monitoring sites were compared to mean background concentrations to assess whether emissions contribute significantly to environmental concentrations. To assess the health significance of potential exposure at nearby residential locations, the 50th percentile concentration was added to the 50th percentile endogenous equivalent concentration and compared to the total equivalent concentration distribution for the non-smoking U.S.

Population: The findings demonstrate that impacts from nearby emission sources are small compared to mean background concentrations at nearby locations, and the total equivalent concentrations for exposed populations are generally indistinguishable from that of the 50th percentile for the non-smoking U.S.

Keywords: contextualization; endogenous equivalent concentration; ethylene oxide; exposure metrics; exposure science; modeling; monitoring; total equivalent concentration.

Figure 1

(a). Modeled (red) and monitored (yellow) sites around Becton Dickinson (Covington, Georgia). (b). Modeled (red) and monitored (yellow) sites around Sterigenics (Smyrna, Georgia). (c). Modeled (red) and monitored (yellow) sites around Sterilization Services (Atlanta, Georgia); sample site F4, which is not shown, is located more than ~5000 m northwest of Sterilization Services.

Figure 1

(a). Modeled (red) and monitored (yellow) sites around Becton Dickinson (Covington, Georgia). (b). Modeled (red) and monitored (yellow) sites around Sterigenics (Smyrna, Georgia). (c). Modeled (red) and monitored (yellow) sites around Sterilization Services (Atlanta, Georgia); sample site F4, which is not shown, is located more than ~5000 m northwest of Sterilization Services.

Figure 2

Adjusted EO concentrations at modeled sites (red: modeled + mean background, 0.18 ppb) and measured EO concentrations at monitored sites (yellow) around Sterilization Services (Atlanta, Georgia).

Figure 3

Highest modeled 5-year average EO concentration for all residential receptors by facility (in absence of background contribution) relative to the mean background EO concentration for Georgia (0.18 ppb). One facility (Sterigenics) had modeling data corresponding to two different time periods, which were delineated by the addition of supplemental EO emission controls.

Figure 4

Estimated total equivalent exposure for the highest 5-year average modeled EO concentration for all residential receptors by facility ((50th percentile endogenous equivalent for the non-smoking U.S. population, or 2.3 ppb) + (50th percentile background EO concentration for Georgia, or 0.12 ppb) + (highest 5-year average modeled EO concentration for all residential receptors by facility)) relative to that of the 50th, 60th, and 95th percentiles of the non-smoking U.S. population (2.5, 2.7, and 5.5 ppb, respectively).