The House Observations of Microbial and Environmental Chemistry (HOMEChem) study was a large-scale collaborative experimental investigation probing indoor air composition and chemistry.

We report elevated levels of gaseous inorganic chlorinated and nitrogenated compounds in indoor air while cleaning with a commercial bleach solution during the House Observations of Microbial and Environmental Chemistry field campaign in summer 2018. Hypochlorous acid (HOCl), chlorine (Cl2), and nitryl chloride (ClNO2) reached part-per-billion by volume levels indoors during bleach cleaningseveral orders of magnitude higher than typically measured in the outdoor atmosphere. Kinetic modeling revealed that multiphase chemistry plays a central role in controlling indoor chlorine and reactive nitrogen chemistry during these periods. Cl2 production occurred via heterogeneous reactions of HOCl on indoor surfaces. ClNO2 and chloramine (NH2Cl, NHCl2, NCl3) production occurred in the applied bleach via aqueous reactions involving nitrite (NO2 –) and ammonia (NH3), respectively. Aqueous-phase and surface chemistry resulted in elevated levels of gas-phase nitrogen dioxide (NO2). We predict hydroxyl (OH) and chlorine (Cl) radical production during these periods (106 and 107 molecules cm–3 s–1, respectively) driven by HOCl and Cl2 photolysis. Ventilation and photolysis accounted for <50% and <0.1% total loss of bleach-related compounds from indoor air, respectively; we conclude that uptake to indoor surfaces is an important additional loss process. Indoor HOCl and nitrogen trichloride (NCl3) mixing ratios during bleach cleaning reported herein are likely detrimental to human health.

A gridded inventory for emissions of methane, ethane, propane, and butanes from oil and gas sources in the Barnett Shale production region has been developed. This inventory extends previous spatially resolved inventories of emissions by characterizing the overall variability in emission magnitudes and the composition of emissions at an hourly time resolution. The inventory is divided into continuous and intermittent emission sources. Sources are defined as continuous if hourly averaged emissions are greater than zero in every hour; otherwise, they are classified as intermittent. In the Barnett Shale, intermittent sources accounted for 14-30% of the mean emissions for methane and 10-34% for ethane, leading to spatial and temporal variability in the location of hourly emissions. The combined variability due to intermittent sources and variability in emission factors can lead to wide confidence intervals in the magnitude and composition of time and location-specific emission inventories; therefore, including temporal and spatial variability in emission inventories is important when reconciling inventories and observations. Comparisons of individual aircraft measurement flights conducted in the Barnett Shale region versus the estimated emission rates for each flight from the emission inventory indicate agreement within the expected variability of the emission inventory for all flights for methane and for all but one flight for ethane.

Spatially resolved emission inventories were used with an atmospheric dispersion model to predict ambient concentrations of methane, ethane, and propane in the Eagle Ford oil and gas production region in south central Texas; predicted concentrations were compared to ground level observations. Using a base case inventory, predicted median propane/ethane concentration ratios were 106% higher (95% CI: 83% higher–226% higher) than observations, while median ethane/methane concentration ratios were 112% higher (95% CI: 17% higher–228% higher) than observations. Predicted median propane and ethane concentrations were factors of 6.9 (95% CI: 3–15.2) and 3.4 (95% CI: 1.4–9) larger than observations, respectively. Predicted median methane concentrations were 7% higher (95% CI: 39% lower–37% higher) than observations. These comparisons indicate that sources of emissions with high propane/ethane ratios (condensate tank flashing) were likely overestimated in the inventories. Because sources of propane and ethane emissions are also sources of methane emissions, the results also suggest that sources of emissions with low ethane/methane ratios (midstream sources) were underestimated. This analysis demonstrates the value of using multiple light alkanes in attributing sources of methane emissions and evaluating the performance of methane emission inventories for oil and natural gas production regions.

No abstract